CN111121943A - Zero point fault detection method and device, computer equipment and readable storage medium - Google Patents
Zero point fault detection method and device, computer equipment and readable storage medium Download PDFInfo
- Publication number
- CN111121943A CN111121943A CN201911284363.1A CN201911284363A CN111121943A CN 111121943 A CN111121943 A CN 111121943A CN 201911284363 A CN201911284363 A CN 201911284363A CN 111121943 A CN111121943 A CN 111121943A
- Authority
- CN
- China
- Prior art keywords
- statistic
- control limit
- hotelling
- state
- prediction error
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000001514 detection method Methods 0.000 title claims abstract description 91
- 238000003860 storage Methods 0.000 title claims abstract description 15
- 238000005303 weighing Methods 0.000 claims abstract description 72
- 238000000034 method Methods 0.000 claims abstract description 27
- 230000001419 dependent effect Effects 0.000 claims description 9
- 208000017604 Hodgkin disease Diseases 0.000 claims description 6
- 238000004590 computer program Methods 0.000 claims description 6
- 239000011159 matrix material Substances 0.000 description 44
- 238000000513 principal component analysis Methods 0.000 description 8
- 238000005070 sampling Methods 0.000 description 8
- 238000011156 evaluation Methods 0.000 description 7
- 238000012360 testing method Methods 0.000 description 6
- 238000004891 communication Methods 0.000 description 4
- 238000000354 decomposition reaction Methods 0.000 description 4
- 238000010276 construction Methods 0.000 description 3
- 238000012937 correction Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 238000009826 distribution Methods 0.000 description 3
- 230000000737 periodic effect Effects 0.000 description 3
- 238000007796 conventional method Methods 0.000 description 2
- 230000001186 cumulative effect Effects 0.000 description 2
- 238000013210 evaluation model Methods 0.000 description 2
- 241001391944 Commicarpus scandens Species 0.000 description 1
- 230000003044 adaptive effect Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000004422 calculation algorithm Methods 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000003745 diagnosis Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
- 238000012847 principal component analysis method Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000012549 training Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01G—WEIGHING
- G01G23/00—Auxiliary devices for weighing apparatus
- G01G23/01—Testing or calibrating of weighing apparatus
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/18—Complex mathematical operations for evaluating statistical data, e.g. average values, frequency distributions, probability functions, regression analysis
Landscapes
- Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Mathematical Optimization (AREA)
- Pure & Applied Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Mathematical Physics (AREA)
- Computational Mathematics (AREA)
- Mathematical Analysis (AREA)
- Life Sciences & Earth Sciences (AREA)
- Operations Research (AREA)
- Probability & Statistics with Applications (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Algebra (AREA)
- Evolutionary Biology (AREA)
- Databases & Information Systems (AREA)
- Software Systems (AREA)
- General Engineering & Computer Science (AREA)
- Bioinformatics & Computational Biology (AREA)
- Testing And Monitoring For Control Systems (AREA)
Abstract
The embodiment of the application provides a zero fault detection method, a zero fault detection device, computer equipment and a readable storage medium, and relates to the technical field of truck scale fault detection, wherein the method comprises the following steps: acquiring state statistics measured by a weighing sensor to be detected; inputting the state statistic into a pre-constructed detection model, and judging whether the state statistic is within a preset control limit range of the state statistic; if so, judging that the weighing sensor to be detected has a fault, and alarming; if not, judging that the weighing sensor to be detected is normal, updating parameters of a pre-constructed detection model according to the state statistics, and accurately detecting the zero point fault of the weighing sensor of the truck scale.
Description
Technical Field
The application relates to the technical field of truck scale fault detection, in particular to a zero fault detection method, a zero fault detection device, computer equipment and a readable storage medium.
Background
The zero fault is one of common fault types of the weighing sensor of the automobile scale, mainly comprises zero fault, zero drift fault, zero impact and periodic interference, belongs to typical tiny faults and is difficult to detect. In actual engineering, a common fault detection method for a truck scale weighing sensor is a manual detection method, namely fault detection is completed by manually loading weights and experience knowledge. The method for detecting the zero fault in the prior art has the advantages of low efficiency, high labor consumption, poor accuracy and difficulty in finding the tiny fault.
In view of this, it is necessary for those skilled in the art to provide a more accurate zero fault detection scheme.
Disclosure of Invention
The embodiment of the application provides a zero fault detection method, a zero fault detection device, computer equipment and a readable storage medium.
The embodiment of the application can be realized as follows:
in a first aspect, an embodiment provides a zero fault detection method, including:
acquiring state statistics measured by a weighing sensor to be detected;
inputting the state statistic into a pre-constructed detection model, and judging whether the state statistic is within a preset control limit range of the state statistic;
if so, judging that the weighing sensor to be detected has a fault, and alarming;
and if not, judging that the weighing sensor to be detected is normal, and updating the parameters of the pre-constructed detection model according to the state statistic.
In an alternative embodiment, the state statistics include a squared prediction error, a hotelling statistic, and a principal component dependent variable residual statistic;
the step of judging whether the state statistic is in the preset control limit range of the state statistic comprises the following steps:
judging whether the state statistic is within a preset control limit range of the state statistic or not according to the square prediction error, the Hotelling statistic, the Hodgkin statistic and the principal element related variable residual statistic;
when the square prediction error is within a first square prediction error control limit range, the Hotelling statistic is within a first Hotelling statistic control limit range, and the Hotelling statistic is within a preset Hotelling statistic constraint range or the principal element related variable residual statistic is within a preset principal element related variable residual statistic constraint range, judging that the state statistic is within the preset control limit range of the state statistic;
or when the square prediction error is in a first square prediction error control limit range, the Hotelling statistic is in a second Hotelling statistic control limit range, the Hotelling statistic is in a preset Hotelling statistic constraint range and the principal component related variable residual statistic is in a preset principal component related variable residual statistic constraint range, the state statistic is judged to be in the preset control limit range of the state statistic;
or when the square prediction error is in a second square prediction error control limit range, the Hotelling statistic is in a first Hotelling statistic control limit range, the Hotelling statistic is in a preset Hotelling statistic constraint range and the principal component related variable residual statistic is in a preset principal component related variable residual statistic constraint range, the state statistic is judged to be in the preset control limit range of the state statistic.
In an alternative embodiment, the first square prediction error control limit range is a control limit that the square prediction error exceeds the square prediction error, the first hotelling statistical quantity control limit range is a control limit that the hotelling statistical quantity exceeds the hotelling statistical quantity, the hotelling statistical quantity preset constraint range is a control limit that the hotelling statistical quantity exceeds the hotelling statistical quantity, and the principal component related variable residual error statistical quantity preset constraint range is a control limit that the principal component related variable residual error statistical quantity exceeds the principal component related variable residual error statistical quantity;
when the square prediction error is within a first square prediction error control limit range, the Hotelling statistic is within a first Hotelling statistic control limit range, and the Hotelling statistic is within a preset Hotelling statistic constraint range or the principal component related variable residual statistic is within a preset principal component related variable residual statistic constraint range, the step of judging that the state statistic is within the preset control limit range of the state statistic comprises the following steps:
and when the square prediction error exceeds the control limit of the square prediction error, the Hotelling statistic exceeds the control limit of the Hotelling statistic, and the Hotelling statistic exceeds the control limit of the Hotelling statistic or the principal component related variable residual error statistic exceeds the control limit of the principal component related variable residual error statistic, judging that the state statistic is in the preset control limit range of the state statistic.
In an alternative embodiment, the second hotelling statistic control limit range is that the hotelling statistic does not exceed the control limit of the hotelling statistic;
when the square prediction error is within a first square prediction error control limit range, the Hotelling statistic is within a second Hotelling statistic control limit range, the Hotelling statistic is within a preset Hotelling statistic constraint range and the principal component related variable residual statistic is within a preset principal component related variable residual statistic constraint range, the step of judging that the state statistic is within the preset control limit range of the state statistic comprises the following steps:
and when the square prediction error exceeds the control limit of the square prediction error, the Hotelling statistic does not exceed the control limit of the Hotelling statistic, the Hotelling statistic exceeds the control limit of the Hotelling statistic, and the principal component related variable residual error statistic exceeds the control limit of the principal component related variable residual error statistic, judging that the state statistic is in the preset control limit range of the state statistic.
In an alternative embodiment, the second square prediction error control limit range is a control limit where the square prediction error does not exceed the square prediction error;
when the square prediction error is within a second square prediction error control limit range, the Hotelling statistic is within a first Hotelling statistic control limit range, the Hotelling statistic is within a preset Hotelling statistic constraint range and the principal component related variable residual statistic is within a preset principal component related variable residual statistic constraint range, the step of judging that the state statistic is within the preset control limit range of the state statistic comprises the following steps:
and when the square prediction error does not exceed the control limit of the square prediction error, the Hotelling statistic exceeds the control limit of the Hotelling statistic, and the principal component related variable residual error statistic exceeds the control limit of the principal component related variable residual error statistic, judging that the state statistic is in the preset control limit range of the state statistic.
In an alternative embodiment, the data is represented by the formula:
judging whether the state statistic is in a preset control limit range of the state statistic, wherein E is a detection model constructed in advance, SPE is a square prediction error, and SPE isαControlling error for square predictionLimit, T2For the Hotelling statistic, TαFor the control limit of the hotelling statistic,
for the Hodgkin's statistic, TH，αFor control limits of Hodgkin's statistics, PVR is principal component-dependent variable residual statistics, PVRαIs the control limit of the principal component related variable residual statistic.
In an alternative embodiment, the method further comprises the step of constructing the pre-constructed detection model, the step comprising:
constructing a recursion principal component model according to the initial principal component model and the original data set;
calculating to obtain the square prediction error, the Hotelling statistic, the Hodgkin statistic and the principal element related variable residual statistic, and the control limit of the square prediction error, the Hotelling statistic, the Hodgkin statistic and the principal element related variable residual statistic according to the recursion principal element model;
and obtaining the pre-constructed detection model according to the square prediction error, the Hotelling statistic and the principal component related variable residual statistic, and the control limits of the square prediction error, the Hotelling statistic and the principal component related variable residual statistic.
In a second aspect, an embodiment provides a zero-point fault detection apparatus, including:
the acquisition module is used for acquiring state statistics measured by the weighing sensor to be detected;
the judging module is used for inputting the state statistic into a pre-constructed detection model and judging whether the state statistic is within a preset control limit range of the state statistic;
if so, judging that the weighing sensor to be detected has a fault, and alarming;
and if not, judging that the weighing sensor to be detected is normal, and updating the parameters of the pre-constructed detection model according to the state statistic.
In a third aspect, embodiments provide a computer device comprising a processor and a non-volatile memory storing computer instructions, wherein when the computer instructions are executed by the processor, the computer device performs the zero-point fault detection method of any one of the preceding embodiments.
In a fourth aspect, an embodiment provides a readable storage medium, which includes a computer program, and the computer program controls a computer device in which the readable storage medium is executed to perform the zero point fault detection method described in any one of the foregoing embodiments.
The beneficial effects of the embodiment of the application include, for example:
by adopting the zero-point fault detection method, the zero-point fault detection device, the computer equipment and the readable storage medium provided by the embodiment of the application, the state statistic measured by the weighing sensor to be detected is obtained, the obtained state statistic is input into the pre-constructed detection model, whether the state statistic is in the preset control limit range of the state statistic is skillfully judged, if yes, the weighing sensor to be detected is judged to have a fault and an alarm is given, if not, the weighing sensor to be detected is judged to be normal, and the pre-constructed detection model is subjected to parameter updating according to the state statistic, so that whether the weighing sensor to be detected has a zero-point fault can be accurately detected.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings that are required to be used in the embodiments will be briefly described below, it should be understood that the following drawings only illustrate some embodiments of the present application and therefore should not be considered as limiting the scope, and for those skilled in the art, other related drawings can be obtained from the drawings without inventive effort.
Fig. 1 is a schematic flowchart illustrating steps of a zero fault detection method according to an embodiment of the present disclosure;
fig. 2 is a schematic flowchart illustrating steps of a detection model construction method according to an embodiment of the present application;
fig. 3 is a schematic block diagram of a structure of a zero fault detection apparatus provided in an embodiment of the present application;
fig. 4 is a block diagram schematically illustrating a structure of a computer device according to an embodiment of the present disclosure.
Icon: 100-a computer device; 110-zero fault detection means; 1101-an acquisition module; 1102-a judgment module; 1103-a building block; 111-a memory; 112-a processor; 113-communication unit.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is obvious that the described embodiments are some embodiments of the present application, but not all embodiments. The components of the embodiments of the present application, generally described and illustrated in the figures herein, can be arranged and designed in a wide variety of different configurations.
Thus, the following detailed description of the embodiments of the present application, presented in the accompanying drawings, is not intended to limit the scope of the claimed application, but is merely representative of selected embodiments of the application. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present application.
It should be noted that: like reference numbers and letters refer to like items in the following figures, and thus, once an item is defined in one figure, it need not be further defined and explained in subsequent figures.
Furthermore, the appearances of the terms "first," "second," and the like, if any, are used solely to distinguish one from another and are not to be construed as indicating or implying relative importance.
It should be noted that the features of the embodiments of the present application may be combined with each other without conflict.
At present, the truck scale is widely applied to the large-scale mass weighing field such as mine ports, transportation and the like, and is a key device for modern logistics weighing and metering. The weighing sensor is a key component of the truck scale, and accumulates output signals of multiple weighing sensors in a parallel connection mode to form a typical multi-sensor system, obtain a voltage signal proportional to the mass of a load to be measured, and transmit the voltage signal to the weighing instrument to finish weighing the load to be measured. The weighing sensor is easy to break down due to severe working environment and working condition change. The zero fault is one of common fault types of the weighing sensor of the automobile scale, mainly comprises zero fault, zero drift fault, zero impact and periodic interference, is a typical tiny fault and is difficult to detect. In actual engineering, a common fault detection method for a truck scale weighing sensor is a manual detection method, namely fault detection is completed by manually loading weights and experience knowledge. The method has the advantages of low efficiency, high labor consumption, poor accuracy and difficulty in finding tiny faults. With the development of fault diagnosis technology, intelligent fault detection methods for weighing sensors are continuously available. Meanwhile, in the existing fault detection method based on principal component analysis, the principal component model does not have self-adaptive capacity, and the requirement of on-line detection of faults of the weighing sensor cannot be met. In addition, the fault detection by using the single statistic has certain limitation and higher false alarm rate. Based on this, the embodiment of the present application provides an example of a zero-point fault detection method, as shown in fig. 1, which includes steps S201 to S204.
Step S201, state statistics measured by a weighing sensor to be detected is obtained.
Step S202, inputting the state statistic into a pre-constructed detection model, and judging whether the state statistic is within a preset control limit range of the state statistic.
If yes, go to step S203.
If not, go to step S204.
And S203, judging that the weighing sensor to be detected has a fault, and alarming.
And S204, judging that the weighing sensor to be detected is normal, and updating parameters of the pre-constructed detection model according to the state statistic.
The pre-constructed detection model can be constructed based on recursive principal component analysis and comprehensive evaluation model, after whether the weighing sensor to be detected fails is judged according to whether the state statistic measured by the weighing sensor to be detected is in the preset control limit range of the state statistic, when the weighing sensor to be detected fails, an alarm can be given, and when the weighing sensor to be detected does not fail, the state statistic obtained at this time can be used as a primary correction value for updating the relevant parameters of the pre-constructed detection model, so that the pre-constructed detection model can be self-updated, and a more accurate detection model can be obtained.
On the basis, the state statistics comprise square prediction error, Hotelling statistics, Hodgkin statistics and principal component related variable residual statistics; the embodiment of the application provides an example for judging whether the state statistic is within the preset control limit range of the state statistic, and the method can be realized through the following steps.
And judging whether the state statistic is in a preset control limit range of the state statistic or not according to the square prediction error, the Hotelling statistic, the Hodgkin statistic and the principal element related variable residual statistic.
And when the square prediction error is within a first square prediction error control limit range, the Hotelling statistic is within a first Hotelling statistic control limit range, and the Hotelling statistic is within a preset Hotelling statistic constraint range or the principal element related variable residual statistic is within a preset principal element related variable residual statistic constraint range, judging that the state statistic is within the preset control limit range of the state statistic.
Or when the square prediction error is in a first square prediction error control limit range, the Hotelling statistic is in a second Hotelling statistic control limit range, the Hotelling statistic is in a preset Hotelling statistic constraint range and the principal component related variable residual statistic is in a preset principal component related variable residual statistic constraint range, the state statistic is judged to be in the preset control limit range of the state statistic.
Or when the square prediction error is in a second square prediction error control limit range, the Hotelling statistic is in a first Hotelling statistic control limit range, the Hotelling statistic is in a preset Hotelling statistic constraint range and the principal component related variable residual statistic is in a preset principal component related variable residual statistic constraint range, the state statistic is judged to be in the preset control limit range of the state statistic.
The square prediction error SPE and the Hotelling statistic T can be used2(i.e. Hotelling's T2) As a basis for detecting the zero point fault, the following may be the detection result: SPE statistics and T2The statistics exceed respective control limits, T2The statistic exceeds the control limit, the SPE statistic does not exceed the control limit, the SPE statistic exceeds the control limit, and T2Statistics not exceeding control limits, SPE statistics and T2None of the statistics exceeds the respective control limit. Wherein, when SPE statistic and T2When the statistics exceed the respective control limits, the weighing sensor to be detected can be considered to be in fault, but the SPE statistics and the T are used2The statistics are all mistakenly alarmed, minor faults (namely the zero fault, the zero drift fault, the zero impact and the periodic interference) are easily ignored, and the Hodgkin statistics
(i.e., Hwakins)
) Sensitive to minor faults, the principal component related variable residual error statistic PVR can better judge the relevance faults of the sensor, so that the method can be utilized
And (4) carrying out constraint on the statistic and the PVR statistic to reduce false alarm. When T is2The statistic exceeds the control limit, the SPE statistic does not exceed the control limit, and only according to T2Whether the weighing sensor to be detected is out or not cannot be judged according to the statistic and the PE statisticThe main reason that the SPE statistic does not exceed the control limit under the condition is that the SPE statistic aims at all sensor variables and has larger conservatism, and the PVR statistic can reflect the change of variables which are obviously related to principal elements in the SPE statistic, so that the detection accuracy can be improved, and meanwhile, the detection accuracy can be improved
Statistics may enhance detection of minor faults while utilizing
And PVR statistics are restricted, so that false alarm and false alarm can be reduced, and whether the weighing sensor to be detected breaks down or not can be determined. When the PE statistics all exceed the control limit, T2The statistic does not exceed the control limit, in which case T occurs2The statistics do not exceed the control limit, the main reason being T caused by some minor faults2The variance of the statistic is not obvious, so that T2The statistics of false alarm can be utilized
Statistics enhance the detection of minor faults, but
The statistics are susceptible to noise, generate a large number of false alarms, and may be constrained using the lower PVR statistics of false alarms. When SPE statistics and T2And the statistics do not exceed respective control limits, and the weighing sensor to be detected can be considered to be not in fault.
On the basis, the first square prediction error control limit range is a control limit that the square prediction error exceeds the square prediction error, the first hotelling statistic control limit range is a control limit that the hotelling statistic exceeds the hotelling statistic, the hotelling statistic preset constraint range is a control limit that the hotelling statistic exceeds the hotelling statistic, and the principal component related variable residual error statistic preset constraint range is a control limit that the principal component related variable residual error statistic exceeds the principal component related variable residual error statistic. The embodiment of the application provides an example of judging that the state statistic is in the preset control limit range of the state statistic when the square prediction error is in the first square prediction error control limit range, the Hotelling statistic is in the first Hotelling statistic control limit range, and the Hotelling statistic is in the preset Hotelling statistic constraint range or the principal component related variable residual statistic is in the preset principal component related variable residual statistic constraint range.
And when the square prediction error exceeds the control limit of the square prediction error, the Hotelling statistic exceeds the control limit of the Hotelling statistic, and the Hotelling statistic exceeds the control limit of the Hotelling statistic or the principal component related variable residual error statistic exceeds the control limit of the principal component related variable residual error statistic, judging that the state statistic is in the preset control limit range of the state statistic.
Specifically, when the SPE statistic exceeds the control limit of the SPE statistic, T2Statistics exceed the T2Can be constrained, i.e. when said
Statistics exceed the
And after the control limit of the statistic or the PVR exceeds the control limit of the PVR and the condition is met, judging that the state statistic is in the preset control limit range of the state statistic.
On the basis of the foregoing, the second hotelling statistic control limit range is that the hotelling statistic does not exceed the control limit of the hotelling statistic. The embodiment of the application provides an example of judging that the state statistic is in the preset control limit range of the state statistic when the square prediction error is in the first square prediction error control limit range, the Hotelling statistic is in the second Hotelling statistic control limit range, the Hotelling statistic is in the preset Hotelling statistic constraint range and the principal component related variable residual statistic is in the preset principal component related variable residual statistic constraint range, and the following steps can be implemented.
And when the square prediction error exceeds the control limit of the square prediction error, the Hotelling statistic does not exceed the control limit of the Hotelling statistic, the Hotelling statistic exceeds the control limit of the Hotelling statistic, and the principal component related variable residual error statistic exceeds the control limit of the principal component related variable residual error statistic, judging that the state statistic is in the preset control limit range of the state statistic.
Specifically, when the SPE statistic exceeds the control limit of the SPE statistic, T2Statistics do not exceed the T2Can be constrained, i.e. when said
Statistics exceed the
And after the control limit of the statistic and the control limit of the PVR are exceeded, the state statistic can be judged to be in the preset control limit range of the state statistic after the four conditions are met.
On this basis, the second square prediction error control limit range is a control limit in which the square prediction error does not exceed the square prediction error. The embodiment of the application provides an example of judging that the state statistic is in the preset control limit range of the state statistic when the square prediction error is in the second square prediction error control limit range, the hotelling statistic is in the first hotelling statistic control limit range, the hotelling statistic is in the preset hotelling statistic constraint range and the principal component related variable residual statistic is in the preset principal component related variable residual statistic constraint range, and the following steps can be implemented.
And when the square prediction error does not exceed the control limit of the square prediction error, the Hotelling statistic exceeds the control limit of the Hotelling statistic, and the principal component related variable residual error statistic exceeds the control limit of the principal component related variable residual error statistic, judging that the state statistic is in the preset control limit range of the state statistic.
Specifically, when the SPE statistic does not exceed the control limit of the SPE statistic, T2Statistics exceed the T2Can be constrained, i.e. when said
Statistics exceed the
And after the control limit of the statistic and the control limit of the PVR are exceeded, the state statistic can be judged to be in the preset control limit range of the state statistic after the four conditions are met.
On the basis of the foregoing, it is possible to obtain the following formula:
judging whether the state statistic is in a preset control limit range of the state statistic, wherein E is a detection model constructed in advance, SPE is a square prediction error, and SPE isαFor controlling the square prediction error, T2For the Hotelling statistic, TαFor the control limit of the hotelling statistic,
for the Hodgkin's statistic, TH，αFor control limits of Hodgkin's statistics, PVR is principal component-dependent variable residual statistics, PVRαIs the control limit of the principal component related variable residual statistic.
On this basis, the embodiment of the present application further provides an example of constructing the pre-constructed detection model, as shown in fig. 2, which can be implemented through step S205 to step S207.
And S205, constructing a recursion principal component model according to the initial principal component model and the original data set.
And step S206, calculating to obtain the square prediction error, the Hotelling statistic, the Hodgkin statistic and the principal element related variable residual statistic, and the control limit of the square prediction error, the Hotelling statistic, the Hodgkin statistic and the principal element related variable residual statistic according to the recursion principal element model.
And step S207, obtaining the pre-constructed detection model according to the square prediction error, the Hotelling statistic and the principal component related variable residual statistic, and the control limits of the square prediction error, the Hotelling statistic and the principal component related variable residual statistic.
The initial principal component model can be constructed firstly, and when the initial principal component model is constructed, sample data can be collected, namely a data set of a plurality of weighing sensors of the truck scale in a no-load state under a normal state is collected, the data set is enough and contains normal fluctuation data of the weighing sensors, so that the principal component model constructed by using the sample data can accurately reflect the normal fault condition of the truck scale, and an original data matrix can be obtained:
wherein,
is a raw data matrix, n is the number of weighing sensors, m is the sampling times,
and sampling data for n paths of weighing sensors for m times.
To reduce the imbalance caused by data size differences, the raw data matrix may be used
Normalizing, i.e. subtracting the corresponding mean value from the data in the matrix, and then dividing by the standard deviation of the data, the mean value of the ith weighing sensor
Expressed by the following formula:
wherein,
mean value of load cell:
variance (variance)
Expressed by the following formula:
wherein,
variance for the load cell:
normalized covariance matrix SmCan be expressed by the following formulas, respectively:
wherein S ismIs a target covariance matrix, X, of the weighing cellmTo normalize the raw data matrix.
For covariance matrix SmOrthogonal decomposition is carried out to obtain principal component variables, namely, the decomposed characteristic values are arranged according to a descending order, the number k of the principal components is determined by a cumulative variance contribution rate method, and when the sum of the current k characteristic values accounts for more than 85% of the total proportion, the corresponding k value is the number of the principal componentsAnd (4) counting.
Thus, the initial principal component vector matrix TmExpressed by the following formula:
Tm＝XmPk
wherein, PkIs a principal component eigenvector matrix.
After the initial principal component model is constructed, the principal component model can be updated by using a recursion algorithm and newly acquired data (which may be standard data specially used for training or state statistics when the weighing sensor to be detected is judged to be normal in the scheme), so that the principal component model is more adaptive to the change of the working condition of the truck scale, namely, on the basis of an established Principal Component Analysis (PCA) model (namely the initial principal component model), the mean value of the ith sensor can be completed by using a Recursive Principal Component Analysis (RPCA) method
Variance (variance)
Covariance matrix SmAnd a feature vector matrix PkOnline recursive computation of. If the length of the data matrix after m times of sampling is m, the length of the matrix containing new sampling data after m +1 times of sampling is m +1, and the mean value and variance of the weighing data of the sensor i are respectively as follows:
wherein,
is the mean value, x, of the weighing data of the weighing sensor i in the raw data matrixi，m+1For the m +1 th sample data of sensor i,
based on this, the normalized data matrix after the m +1 th sampling can be expressed as:
wherein,
as a mean matrix:
diag(σ1，m+1，σ2，m+1，...，σn，m+1) As a standard matrix, the normalized weighing data matrix of the m +1 th sample can be represented by the following formula:
therein, sigmam+1＝diag(σ1，m+1，σ2，m+1，...，σn，m+1) Is a matrix of standard deviations of the weighing data,
for m +1 sampled weighing data, the updated covariance matrix can be represented by the following formula:
wherein,
because recursion is carried out once per sampling, the total variance variation of the input data matrix is very small, and the updated covariance matrix Sm+1It can also be expressed by the following formula:
then to Sm+1After orthogonal decomposition, principal component update can be performed by using rank 1 correction of the matrix, namely, newly acquired data is used
For covariance matrix SmAnd obtaining a recursion principal component model in the process of correction.
Specifically, P can be setmAnd OmAre respectively covariance matrix SmThe feature vector matrix and the diagonal matrix obtained after orthogonal decomposition are the corrected covariance matrix Sm+1Comprises the following steps:
wherein,
in the form of a diagonal matrix,
D+σuuTin which both the newly sampled weighing data vector x is includedm+1And also includes the last calculation result PmAnd Om. For D + sigma uuTAfter orthogonal decomposition, we can get:
D+σuuT＝QθQT
based on this, the updated eigenvector matrix is Pm+1＝PmQ, a diagonal matrix O with characteristic values as diagonal elements after updatingm+1And if theta is equal to theta, the updated principal component vector matrix is as follows:
wherein,
the updated principal component eigenvector matrix can be derived from the eigenvector matrix P by the cumulative variance contribution ratio method as described abovem+1Determining principal component eigenvector matrix
The sum of the first k characteristic values here accounts for a proportion of more than 85% of the sum.
Obtaining updated principal component vector matrix Tm+1On the basis, the current output number x ═ x of the weighing sensor can be obtained by using a recursion principal component analysis method1，x2，...，xm+1]The estimated values of (c) are:
based on this, the difference (i.e. residual) between the actual sampling value and the estimated value of the weighing sensor can be calculated as:
wherein I is an identity matrix. Under normal conditions, Δ x is mainly measurement noise, and under fault conditions, the value of Δ x will be increased significantly, and the fault detection of the sensor can be realized by using related detection indexes.
Obtaining updated principal component vector matrix Tm+1Based on the sum residual Δ x, Hotelling's T2Statistical sum Hotelling's T2Control limit of statistic TαCalculated by the following formula:
wherein A is1，k＝diag{λ1，λ2，...，λnThe method comprises the following steps of (1) taking an eigenvalue matrix as an index, and taking x as an actual output value of a weighing sensor:
where m and l are degrees of freedom, α is the confidence, Fl，m-l，αThe F distribution with the degrees of freedom of l and m-l and the confidence of α is a critical value, and α is generally 85% considering the influence of the false alarm rate.
SPE statistics and control limit SPE of SPEαCalculated by the following formula:
wherein, Δ x is the residual error of the weighing data, I is a unit matrix, and x is the actual output value of the weighing sensor;
wherein,
α are used as the confidence level,
Cαfor a threshold of α confidence in a standard normal distribution, α is typically 85% or 90%.
wherein, Λk+1，n＝diag{λ1，λ2，...，λnIs a matrix of eigenvalues,
the scoring matrix is composed of the last (n-k) eigenvectors, and x is the actual output value of the weighing sensor;
where m, n, and k are degrees of freedom, α is confidence, Fn-k，m-n+k，αThe F distribution with degrees of freedom n-k and m-n + k and a confidence of α is critical.
PVR statistics and PVR limit control for PVR statisticsαCalculated by the following formula:
wherein,
is a feature matrix corresponding to the first s pivot variables. Generally, when the percentage of the complex correlation coefficient of the current s pivot element variables in the total number is more than 85%, the corresponding value of s is the result, which is the output value of the estimator formed by the most relevant s pivot elements, s is less than k, and k is the number of elements;
PVRα＝w SPEα
where w is the percentage of the first s pivot variables to the total variables, i.e.
Wherein γ is a complex correlation coefficient of an independent variable and a principal component variable, i.e.
By Hotelling's T described above2Statistics, SPE statistics,
Statistics and PVR statistics and respective corresponding controlsLimit Tα、SPEα、TH，αAnd SPEαThe previously constructed detection model E can be constructed. The overall false alarm rate of the zero fault detection method based on the recursion principal component analysis and comprehensive evaluation model provided by the embodiment of the application is reduced by one order of magnitude compared with the false alarm rate detected by the traditional method, the fault detection range is obviously enlarged, the zero fault and the micro fault can be well detected when occurring, and the real-time performance and the reliability of the weighing system are improved.
Referring to table one, the embodiment of the present application provides an example of testing for zero faults. 500 groups (500 x 6) of zero data under the normal working condition of the truck scale can be collected as an original data set, an initial PCA model is established, and preparation is made for online detection; then respectively utilize T2The method,
The method, the SPE method, the PVR method and the comprehensive evaluation method are used for testing zero faults on line and testing 3500 groups of data in total. (some of these data contain zero point failures of different intensities). And the first table is a zero fault detection result. In table one, 0, 0.1, and 0.2 indicate that the failure strengths are 0, 0.1, and 0.2, respectively. The table shows that the false alarm rate of the comprehensive evaluation method is far lower than that of the traditional method under the same fault intensity, which indicates that the comprehensive evaluation method has better fault discrimination. Meanwhile, it can be found that the PVR is more easy to miss the alarm,
false alarms are more likely to occur when no faults occur.
Watch 1
The embodiment of the application also provides an example for testing the null shift fault, and the table II can be referred to. Collecting 500 groups (500 x 6) of zero data under the normal working condition of the truck scale as an original data set, establishing an initial PCA model, and making a standard for online detectionPreparing; then respectively utilize T2Method, T2The method, the SPE method, the PVR method and the comprehensive evaluation method are used for testing zero faults on line and testing 3500 groups of data in total. (some of these data contain zero point failures of different intensities). And the second table is the bit failure detection result. In table two, 1, 2, 5, 10, and 20 indicate that the failure strengths are 1, 2, 5, 10, and 20, respectively. It can be seen from the table that when the sensor has a null shift fault, since the null value changes irregularly, when fault detection is performed by using the conventional method and the comprehensive evaluation method proposed herein, false alarms occur, but the false alarm rate of the comprehensive evaluation method is lower than that of the conventional method under the condition that the fault intensities are the same (i.e., the noise intensities are the same), and the false alarm rate of the fault detection is further reduced along with the increase of the null shift fault intensity, which indicates that the comprehensive evaluation method has higher accuracy.
Watch two
The embodiment of the present application provides a zero fault detection apparatus 110, as shown in fig. 3, the apparatus includes:
the obtaining module 1101 is configured to obtain a state statistic measured by the to-be-detected weighing sensor.
The judging module 1102 is configured to input the state statistic into a pre-constructed detection model, and judge whether the state statistic is within a preset control limit range of the state statistic.
And if so, judging that the weighing sensor to be detected has a fault, and alarming.
And if not, judging that the weighing sensor to be detected is normal, and updating the parameters of the pre-constructed detection model according to the state statistic.
Further, the state statistics include a squared prediction error, a hotelling statistic, and a principal component dependent variable residual statistic. The determining module 1102 is specifically configured to:
judging whether the state statistic is within a preset control limit range of the state statistic or not according to the square prediction error, the Hotelling statistic, the Hodgkin statistic and the principal element related variable residual statistic; when the square prediction error is within a first square prediction error control limit range, the Hotelling statistic is within a first Hotelling statistic control limit range, and the Hotelling statistic is within a preset Hotelling statistic constraint range or the principal element related variable residual statistic is within a preset principal element related variable residual statistic constraint range, judging that the state statistic is within the preset control limit range of the state statistic; or when the square prediction error is in a first square prediction error control limit range, the Hotelling statistic is in a second Hotelling statistic control limit range, the Hotelling statistic is in a preset Hotelling statistic constraint range and the principal component related variable residual statistic is in a preset principal component related variable residual statistic constraint range, the state statistic is judged to be in the preset control limit range of the state statistic; or when the square prediction error is in a second square prediction error control limit range, the Hotelling statistic is in a first Hotelling statistic control limit range, the Hotelling statistic is in a preset Hotelling statistic constraint range and the principal component related variable residual statistic is in a preset principal component related variable residual statistic constraint range, the state statistic is judged to be in the preset control limit range of the state statistic.
Further, the first square prediction error control limit range is a control limit where the square prediction error exceeds the square prediction error, the first hotelling statistic control limit range is a control limit where the hotelling statistic exceeds the hotelling statistic, the hotelling statistic preset constraint range is a control limit where the hotelling statistic exceeds the hotelling statistic, and the principal component related variable residual error statistic preset constraint range is a control limit where the principal component related variable residual error statistic exceeds the principal component related variable residual error statistic. The determining module 1102 is further specifically configured to:
and when the square prediction error exceeds the control limit of the square prediction error, the Hotelling statistic exceeds the control limit of the Hotelling statistic, and the Hotelling statistic exceeds the control limit of the Hotelling statistic or the principal component related variable residual error statistic exceeds the control limit of the principal component related variable residual error statistic, judging that the state statistic is in the preset control limit range of the state statistic.
Further, the second hotelling statistic control limit range is that the hotelling statistic does not exceed the control limit of the hotelling statistic. The determining module 1102 is further specifically configured to:
and when the square prediction error exceeds the control limit of the square prediction error, the Hotelling statistic does not exceed the control limit of the Hotelling statistic, the Hotelling statistic exceeds the control limit of the Hotelling statistic, and the principal component related variable residual error statistic exceeds the control limit of the principal component related variable residual error statistic, judging that the state statistic is in the preset control limit range of the state statistic.
Further, the second square prediction error control limit range is a control limit where the square prediction error does not exceed the square prediction error. The determining module 1102 is further specifically configured to:
and when the square prediction error does not exceed the control limit of the square prediction error, the Hotelling statistic exceeds the control limit of the Hotelling statistic, and the principal component related variable residual error statistic exceeds the control limit of the principal component related variable residual error statistic, judging that the state statistic is in the preset control limit range of the state statistic.
Further, the apparatus further comprises a construction module 1103, and the construction module 1103 is configured to:
constructing a recursion principal component model according to the initial principal component model and the original data set; calculating to obtain the square prediction error, the Hotelling statistic, the Hodgkin statistic and the principal element related variable residual statistic, and the control limit of the square prediction error, the Hotelling statistic, the Hodgkin statistic and the principal element related variable residual statistic according to the recursion principal element model; and obtaining the pre-constructed detection model according to the square prediction error, the Hotelling statistic and the principal component related variable residual statistic, and the control limits of the square prediction error, the Hotelling statistic and the principal component related variable residual statistic.
The embodiment of the present application provides a computer device 100, where the computer device 100 includes a processor and a non-volatile memory storing computer instructions, and when the computer instructions are executed by the processor, the computer device 100 executes the aforementioned zero point fault detection method. As shown in fig. 4, fig. 4 is a block diagram of a computer device 100 according to an embodiment of the present disclosure. The computer apparatus 100 includes a zero point fault detection device 110, a memory 111, a processor 112, and a communication unit 113.
The memory 111, the processor 112 and the communication unit 113 are electrically connected to each other directly or indirectly to realize data transmission or interaction. For example, the components may be electrically connected to each other via one or more communication buses or signal lines. The zero-point fault detection apparatus 110 includes at least one software functional module which can be stored in the memory 111 in the form of software or firmware (firmware) or is fixed in an Operating System (OS) of the computer device 100. The processor 112 is used for executing executable modules stored in the memory 111, such as software functional modules and computer programs included in the zero point fault detection apparatus 110.
The Memory 111 may be, but is not limited to, a Random Access Memory (RAM), a Read Only Memory (ROM), a Programmable Read-Only Memory (PROM), an Erasable Read-Only Memory (EPROM), an electrically Erasable Read-Only Memory (EEPROM), and the like.
An embodiment of the present application provides a readable storage medium, where the readable storage medium includes a computer program, and when the computer program runs, the computer apparatus 100 where the readable storage medium is located is controlled to execute the foregoing zero point fault detection method.
In summary, the embodiment of the present application provides a zero point fault detection method, an apparatus, a computer device, and a readable storage medium, where a state statistic measured by a to-be-detected weighing sensor is obtained, the obtained state statistic is input into a pre-constructed detection model, whether the state statistic is within a preset control limit range of the state statistic is skillfully determined, if yes, it is determined that the to-be-detected weighing sensor is faulty, an alarm is issued, if not, it is determined that the to-be-detected weighing sensor is normal, and parameter update is performed on the pre-constructed detection model according to the state statistic, so that whether a zero point fault exists in the to-be-detected weighing sensor can be accurately detected, and self-update of the detection model is achieved.
The above description is only for the specific embodiments of the present application, but the scope of the present application is not limited thereto, and any changes or substitutions that can be easily conceived by those skilled in the art within the technical scope of the present application should be covered within the scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims.
Claims (10)
1. A zero-point fault detection method, comprising:
acquiring state statistics measured by a weighing sensor to be detected;
inputting the state statistic into a pre-constructed detection model, and judging whether the state statistic is within a preset control limit range of the state statistic;
if so, judging that the weighing sensor to be detected has a fault, and alarming;
and if not, judging that the weighing sensor to be detected is normal, and updating the parameters of the pre-constructed detection model according to the state statistic.
2. The method of claim 1, wherein the state statistics include a squared prediction error, a hotelling statistic, and a principal component dependent variable residual statistic;
the step of judging whether the state statistic is in the preset control limit range of the state statistic comprises the following steps:
judging whether the state statistic is within a preset control limit range of the state statistic or not according to the square prediction error, the Hotelling statistic, the Hodgkin statistic and the principal element related variable residual statistic;
when the square prediction error is within a first square prediction error control limit range, the Hotelling statistic is within a first Hotelling statistic control limit range, and the Hotelling statistic is within a preset Hotelling statistic constraint range or the principal element related variable residual statistic is within a preset principal element related variable residual statistic constraint range, judging that the state statistic is within the preset control limit range of the state statistic;
or when the square prediction error is in a first square prediction error control limit range, the Hotelling statistic is in a second Hotelling statistic control limit range, the Hotelling statistic is in a preset Hotelling statistic constraint range and the principal component related variable residual statistic is in a preset principal component related variable residual statistic constraint range, the state statistic is judged to be in the preset control limit range of the state statistic;
or when the square prediction error is in a second square prediction error control limit range, the Hotelling statistic is in a first Hotelling statistic control limit range, the Hotelling statistic is in a preset Hotelling statistic constraint range and the principal component related variable residual statistic is in a preset principal component related variable residual statistic constraint range, the state statistic is judged to be in the preset control limit range of the state statistic.
3. The method of claim 2, wherein the first square prediction error control limit range is a control limit at which the square prediction error exceeds the square prediction error, the first hotelling statistic control limit range is a control limit at which the hotelling statistic exceeds the hotelling statistic, the hotelling statistic preset constraint range is a control limit at which the hotelling statistic exceeds the hotelling statistic, and the principal component dependent variable residual statistic preset constraint range is a control limit at which the principal component dependent variable residual statistic exceeds the principal component dependent variable residual statistic;
when the square prediction error is within a first square prediction error control limit range, the Hotelling statistic is within a first Hotelling statistic control limit range, and the Hotelling statistic is within a preset Hotelling statistic constraint range or the principal component related variable residual statistic is within a preset principal component related variable residual statistic constraint range, the step of judging that the state statistic is within the preset control limit range of the state statistic comprises the following steps:
and when the square prediction error exceeds the control limit of the square prediction error, the Hotelling statistic exceeds the control limit of the Hotelling statistic, and the Hotelling statistic exceeds the control limit of the Hotelling statistic or the principal component related variable residual error statistic exceeds the control limit of the principal component related variable residual error statistic, judging that the state statistic is in the preset control limit range of the state statistic.
4. The method of claim 3, wherein the second range of hotelling statistical quantity control limits is such that the hotelling statistical quantity does not exceed the hotelling statistical quantity control limit;
when the square prediction error is within a first square prediction error control limit range, the Hotelling statistic is within a second Hotelling statistic control limit range, the Hotelling statistic is within a preset Hotelling statistic constraint range and the principal component related variable residual statistic is within a preset principal component related variable residual statistic constraint range, the step of judging that the state statistic is within the preset control limit range of the state statistic comprises the following steps:
and when the square prediction error exceeds the control limit of the square prediction error, the Hotelling statistic does not exceed the control limit of the Hotelling statistic, the Hotelling statistic exceeds the control limit of the Hotelling statistic, and the principal component related variable residual error statistic exceeds the control limit of the principal component related variable residual error statistic, judging that the state statistic is in the preset control limit range of the state statistic.
5. The method of claim 4, wherein the second square prediction error control limit range is a control limit at which the square prediction error does not exceed the square prediction error;
when the square prediction error is within a second square prediction error control limit range, the Hotelling statistic is within a first Hotelling statistic control limit range, the Hotelling statistic is within a preset Hotelling statistic constraint range and the principal component related variable residual statistic is within a preset principal component related variable residual statistic constraint range, the step of judging that the state statistic is within the preset control limit range of the state statistic comprises the following steps:
and when the square prediction error does not exceed the control limit of the square prediction error, the Hotelling statistic exceeds the control limit of the Hotelling statistic, and the principal component related variable residual error statistic exceeds the control limit of the principal component related variable residual error statistic, judging that the state statistic is in the preset control limit range of the state statistic.
6. The method of claim 2, characterized by the formula:
judging whether the state statistic is in a preset control limit range of the state statistic, wherein E is a detection model constructed in advance, SPE is a square prediction error, and SPE isαFor controlling the square prediction error, T2For the Hotelling statistic, TαFor the control limit of the hotelling statistic,
for the Hodgkin's statistic, TH,αFor control limits of Hodgkin's statistics, PVR is principal component-dependent variable residual statistics, PVRαIs the control limit of the principal component related variable residual statistic.
7. The method of claim 2, further comprising the step of constructing the pre-constructed detection model, comprising:
constructing a recursion principal component model according to the initial principal component model and the original data set;
calculating to obtain the square prediction error, the Hotelling statistic, the Hodgkin statistic and the principal element related variable residual statistic, and the control limit of the square prediction error, the Hotelling statistic, the Hodgkin statistic and the principal element related variable residual statistic according to the recursion principal element model;
and obtaining the pre-constructed detection model according to the square prediction error, the Hotelling statistic and the principal component related variable residual statistic, and the control limits of the square prediction error, the Hotelling statistic and the principal component related variable residual statistic.
8. A zero-point fault detection device, comprising:
the acquisition module is used for acquiring state statistics measured by the weighing sensor to be detected;
the judging module is used for inputting the state statistic into a pre-constructed detection model and judging whether the state statistic is within a preset control limit range of the state statistic;
if so, judging that the weighing sensor to be detected has a fault, and alarming;
and if not, judging that the weighing sensor to be detected is normal, and updating the parameters of the pre-constructed detection model according to the state statistic.
9. A computer device comprising a processor and a non-volatile memory having computer instructions stored thereon, wherein the computer instructions, when executed by the processor, cause the computer device to perform the zero-point fault detection method of any one of claims 1-7.
10. A readable storage medium, characterized in that the readable storage medium comprises a computer program which, when running, controls a computer device in which the readable storage medium is located to perform the zero point fault detection method of any one of claims 1-7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911284363.1A CN111121943B (en) | 2019-12-13 | 2019-12-13 | Zero point fault detection method and device, computer equipment and readable storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911284363.1A CN111121943B (en) | 2019-12-13 | 2019-12-13 | Zero point fault detection method and device, computer equipment and readable storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111121943A true CN111121943A (en) | 2020-05-08 |
| CN111121943B CN111121943B (en) | 2021-05-18 |
Family
ID=70498774
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911284363.1A Active CN111121943B (en) | 2019-12-13 | 2019-12-13 | Zero point fault detection method and device, computer equipment and readable storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111121943B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114397003A (en) * | 2021-12-27 | 2022-04-26 | 深圳市晨北科技有限公司 | Weighing calibration method, device, equipment and medium |
| CN115615531A (en) * | 2022-09-26 | 2023-01-17 | 郑州衡量科技股份有限公司 | Zero point correction method, data processing device and readable storage medium |
Citations (20)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN85106408A (en) * | 1985-07-05 | 1986-12-31 | 勒鲁瓦·赫根布 | Device and method for reacting on-board measurement of the load carried by the body of a truck |
| EP0805421A2 (en) * | 1996-05-02 | 1997-11-05 | Francotyp-Postalia AG & Co. | Method and arrangement for processing data in a mail processing system with a franking machine |
| CN1253621A (en) * | 1997-12-26 | 2000-05-17 | 日本碍子株式会社 | Double-headed mass sensor and mass detection method |
| WO2003106710A1 (en) * | 2002-06-18 | 2003-12-24 | P & I Consulting Co., Ltd. | Correction method, apparatus and recording medium on oligonucleotide microarray using principal component analysis |
| US7198964B1 (en) * | 2004-02-03 | 2007-04-03 | Advanced Micro Devices, Inc. | Method and apparatus for detecting faults using principal component analysis parameter groupings |
| CN101042306A (en) * | 2006-12-31 | 2007-09-26 | 李彩珍 | Intelligent digital balance |
| CN101263371A (en) * | 2005-09-15 | 2008-09-10 | 皇家飞利浦电子股份有限公司 | Weight management system with simple data input |
| CN102778355A (en) * | 2012-08-07 | 2012-11-14 | 北京交通大学 | Rolling bearing state identification method based on empirical mode decomposition (EMD) and principal component analysis (PCA) |
| CN103424669A (en) * | 2013-08-05 | 2013-12-04 | 昆明理工大学 | Route selection method for analyzing first principal component by utilizing principal component of zero-sequence current matrix of faulty feeder line |
| CN103472820A (en) * | 2013-09-18 | 2013-12-25 | 哈尔滨工业大学 | Method for diagnosing propulsion system failure based on partial least squares algorithm |
| CN103699714A (en) * | 2013-12-01 | 2014-04-02 | 北京航空航天大学 | Flexible object real-time cutting simulation method based on finite element and meshless coupling |
| US8825586B2 (en) * | 2009-02-24 | 2014-09-02 | International Road Dynamics | Vehicle type recognition at a checkpoint using PCA and BIC |
| CN106197633A (en) * | 2016-07-05 | 2016-12-07 | 厦门市三友软件有限公司 | Vehicle weighing method, vehicle weighing device and vehicle weighing system |
| CN107544477A (en) * | 2017-10-23 | 2018-01-05 | 中国石油大学(华东) | Nonlinear industrial processes fault detection method based on core pivot element analysis |
| US10001402B1 (en) * | 2015-06-19 | 2018-06-19 | Amazon Technologies, Inc. | Instrumented item stowage system with modular elements |
| CN108762228A (en) * | 2018-05-25 | 2018-11-06 | 江南大学 | A kind of multi-state fault monitoring method based on distributed PCA |
| CN109141598A (en) * | 2018-08-20 | 2019-01-04 | 中北大学 | A kind of vehicle dynamic overload monitoring system based on distributed shock sensor array |
| CN109528213A (en) * | 2018-12-24 | 2019-03-29 | 无锡市儿童医院 | Newborn's information collection and identifying system and method |
| CN109782739A (en) * | 2019-01-29 | 2019-05-21 | 中国能源建设集团广东省电力设计研究院有限公司 | Repair method, device, computer equipment and the storage medium of equipment fault |
| CN109799808A (en) * | 2019-01-25 | 2019-05-24 | 东北大学 | A kind of dynamic process failure prediction method based on reconfiguration technique |
-
2019
- 2019-12-13 CN CN201911284363.1A patent/CN111121943B/en active Active
Patent Citations (20)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN85106408A (en) * | 1985-07-05 | 1986-12-31 | 勒鲁瓦·赫根布 | Device and method for reacting on-board measurement of the load carried by the body of a truck |
| EP0805421A2 (en) * | 1996-05-02 | 1997-11-05 | Francotyp-Postalia AG & Co. | Method and arrangement for processing data in a mail processing system with a franking machine |
| CN1253621A (en) * | 1997-12-26 | 2000-05-17 | 日本碍子株式会社 | Double-headed mass sensor and mass detection method |
| WO2003106710A1 (en) * | 2002-06-18 | 2003-12-24 | P & I Consulting Co., Ltd. | Correction method, apparatus and recording medium on oligonucleotide microarray using principal component analysis |
| US7198964B1 (en) * | 2004-02-03 | 2007-04-03 | Advanced Micro Devices, Inc. | Method and apparatus for detecting faults using principal component analysis parameter groupings |
| CN101263371A (en) * | 2005-09-15 | 2008-09-10 | 皇家飞利浦电子股份有限公司 | Weight management system with simple data input |
| CN101042306A (en) * | 2006-12-31 | 2007-09-26 | 李彩珍 | Intelligent digital balance |
| US8825586B2 (en) * | 2009-02-24 | 2014-09-02 | International Road Dynamics | Vehicle type recognition at a checkpoint using PCA and BIC |
| CN102778355A (en) * | 2012-08-07 | 2012-11-14 | 北京交通大学 | Rolling bearing state identification method based on empirical mode decomposition (EMD) and principal component analysis (PCA) |
| CN103424669A (en) * | 2013-08-05 | 2013-12-04 | 昆明理工大学 | Route selection method for analyzing first principal component by utilizing principal component of zero-sequence current matrix of faulty feeder line |
| CN103472820A (en) * | 2013-09-18 | 2013-12-25 | 哈尔滨工业大学 | Method for diagnosing propulsion system failure based on partial least squares algorithm |
| CN103699714A (en) * | 2013-12-01 | 2014-04-02 | 北京航空航天大学 | Flexible object real-time cutting simulation method based on finite element and meshless coupling |
| US10001402B1 (en) * | 2015-06-19 | 2018-06-19 | Amazon Technologies, Inc. | Instrumented item stowage system with modular elements |
| CN106197633A (en) * | 2016-07-05 | 2016-12-07 | 厦门市三友软件有限公司 | Vehicle weighing method, vehicle weighing device and vehicle weighing system |
| CN107544477A (en) * | 2017-10-23 | 2018-01-05 | 中国石油大学(华东) | Nonlinear industrial processes fault detection method based on core pivot element analysis |
| CN108762228A (en) * | 2018-05-25 | 2018-11-06 | 江南大学 | A kind of multi-state fault monitoring method based on distributed PCA |
| CN109141598A (en) * | 2018-08-20 | 2019-01-04 | 中北大学 | A kind of vehicle dynamic overload monitoring system based on distributed shock sensor array |
| CN109528213A (en) * | 2018-12-24 | 2019-03-29 | 无锡市儿童医院 | Newborn's information collection and identifying system and method |
| CN109799808A (en) * | 2019-01-25 | 2019-05-24 | 东北大学 | A kind of dynamic process failure prediction method based on reconfiguration technique |
| CN109782739A (en) * | 2019-01-29 | 2019-05-21 | 中国能源建设集团广东省电力设计研究院有限公司 | Repair method, device, computer equipment and the storage medium of equipment fault |
Non-Patent Citations (9)
| Title |
|---|
| ASGHAR F , TALHA M , KIM S Y , ET AL.: "Hotelling T Index based PCA Method for Fault Detection in Transient State Processes", 《 INSTITUTE OF CONTROL, ROBOTICS AND SYSTEMS》 * |
| 刘洋等: "称重设备传感器零漂检测方法研究", 《微型机与应用》 * |
| 安星等: "基于累积残差贡献率的传感器故障定位方法", 《土木与环境工程学报（中英文）》 * |
| 张铭钧等: "《水下机器人故障诊断理论与技术》", 30 November 2016, 哈尔滨工程大学出版社 * |
| 杜海莲等: "基于主元分析的多传感器故障检测", 《南京航空航天大学学报》 * |
| 林海军 等: "基于信息融合的汽车衡称重传感器故障诊断", 《控制理论与应用》 * |
| 程龙: "改进的递推主元分析及递推主元回归算法", 《控制工程》 * |
| 胡云鹏等: "基于主元分析及递推主元分析的AHU传感器故障检测", 《全国制冷空调新技术研讨会》 * |
| 邱天等: "基于霍金斯指标的传感器数据邮箱性验证", 《中国机电工程学报》 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114397003A (en) * | 2021-12-27 | 2022-04-26 | 深圳市晨北科技有限公司 | Weighing calibration method, device, equipment and medium |
| CN115615531A (en) * | 2022-09-26 | 2023-01-17 | 郑州衡量科技股份有限公司 | Zero point correction method, data processing device and readable storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111121943B (en) | 2021-05-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108052770B (en) | Long-span bridge girder performance early warning method considering time-varying effect | |
| CN110515351B (en) | Abnormality detector | |
| CN111444953B (en) | Sensor fault monitoring method based on improved particle swarm optimization algorithm | |
| Jiang et al. | Weighted kernel principal component analysis based on probability density estimation and moving window and its application in nonlinear chemical process monitoring | |
| KR102116711B1 (en) | Early warning system and method of power plants using multiple prediction model | |
| CN108549908B (en) | Chemical process fault detection method based on multi-sampling probability kernel principal component model | |
| CN103631145B (en) | Multiple operating modes process monitoring method and system based on monitor control index switching | |
| CN107016236A (en) | Power network false data detection method for injection attack based on non-linear measurement equation | |
| CN110909822A (en) | Satellite anomaly detection method based on improved Gaussian process regression model | |
| KR20190025474A (en) | Apparatus and Method for Predicting Plant Data | |
| CN115268417B (en) | Self-adaptive ECU fault diagnosis control method | |
| CN111121943B (en) | Zero point fault detection method and device, computer equipment and readable storage medium | |
| CN107644148A (en) | On-orbit satellite abnormal state monitoring method and system based on multi-parameter association | |
| CN111368428A (en) | Sensor precision reduction fault detection method based on monitoring second-order statistics | |
| CN106767898B (en) | A method of detection measuring system of satellite attitude small fault | |
| CN110308713A (en) | A kind of industrial process failure identification variables method based on k neighbour reconstruct | |
| CN103207567B (en) | Low-false-alarm-rate improved principal component analysis process monitoring method and system | |
| CN110266527B (en) | Sensor node fault classification alarm method and device based on spatial correlation | |
| CN105550457A (en) | Dynamic evolution model correction method and system | |
| Yang et al. | A real-time fault detection and isolation strategy for gas sensor arrays | |
| CN109725138B (en) | Method, device and equipment for detecting quality of clinker in production of intelligent cement factory | |
| US7366639B2 (en) | Methods for establishing alerts and/or alert limits for monitoring mechanical devices | |
| CN112098066A (en) | High-voltage shunt reactor fault diagnosis method and system based on gate control circulation unit | |
| CN109506676A (en) | Earth sensor method for diagnosing faults based on regression modeling | |
| CN115618173A (en) | Geological drilling process anomaly detection method based on interval-augmented Mahalanobis distance |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |