CN110231447A - The method, apparatus and terminal device of water quality abnormality detection - Google Patents
The method, apparatus and terminal device of water quality abnormality detection Download PDFInfo
- Publication number
- CN110231447A CN110231447A CN201910497097.4A CN201910497097A CN110231447A CN 110231447 A CN110231447 A CN 110231447A CN 201910497097 A CN201910497097 A CN 201910497097A CN 110231447 A CN110231447 A CN 110231447A
- Authority
- CN
- China
- Prior art keywords
- data
- duration
- measurement
- residual error
- moment
- 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.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N33/00—Investigating or analysing materials by specific methods not covered by groups G01N1/00 - G01N31/00
- G01N33/18—Water
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A20/00—Water conservation; Efficient water supply; Efficient water use
- Y02A20/20—Controlling water pollution; Waste water treatment
Landscapes
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Chemical & Material Sciences (AREA)
- Engineering & Computer Science (AREA)
- Food Science & Technology (AREA)
- Medicinal Chemistry (AREA)
- Physics & Mathematics (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- General Health & Medical Sciences (AREA)
- General Physics & Mathematics (AREA)
- Immunology (AREA)
- Pathology (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The present invention is suitable for water quality inspection technique field, provide the method, apparatus and terminal device of a kind of water quality abnormality detection, the described method includes: obtaining the time series measurement data in the first duration of water area monitoring node to be measured, time series measurement data is inputted into the UV254 prediction model based on Recognition with Recurrent Neural Network RNN, obtains the UV254 prediction data in first duration of water area monitoring node to be measured；Difference is calculated using UV254 measurement data and UV254 prediction data, obtains residual error, confidence interval is generated according to residual distribution；If residual error is determined as normal point in the confidence interval；If residual error in the confidence interval, is not determined as abnormal point；If the cumulative probability is greater than predetermined probabilities threshold value determine that water quality anomalous event has occurred in second time window for the cumulative probability that water quality exception probability occurs in measurement the second time windows of moment computational representation all in the second time window in the first duration.High water quality abnormality detection effect drops in the present invention significantly.
Description
Technical field
The invention belongs to water quality inspection technique field more particularly to a kind of method, apparatus and terminal of water quality abnormality detection
Equipment.
Background technique
Water resources in china lacks, and water resources ownership per capita only has a quarter of world average level, however China
Part water resource is by different degrees of pollution, and wherein the specific gravity of the organic contamination event in river water quality contamination accident is larger,
Influence urban water supply.UV254 is an important control parameter for measuring Organic substance in water index, and can effectively reflect in water has
The content of machine object, have the characteristics that detect speed it is fast, to organic matter high sensitivity, without reagent, can effectively detect river
In content of organics, therefore be suitable for the water quality abnormality detection in river, in particular for the accurate of organic contamination event and
When detection, ensure urban water supply safety.
In recent years, with the development of machine learning, related algorithm makes great progress in abnormality detection field, especially
It is abnormal water detection method based on data-driven, such as SVM, random forest etc..But these methods are examined extremely in water quality
Effect is bad during survey, and accuracy is not high enough.
Summary of the invention
In view of this, the embodiment of the invention provides the method, apparatus and terminal device of a kind of water quality abnormality detection, with solution
Certainly the relevant technologies effect in water quality abnormality detecting process is bad, the not high enough technical problem of accuracy.
The first aspect of the embodiment of the present invention provides a kind of method of water quality abnormality detection, comprising:
The UV254 measurement data at each measurement moment in the first duration of water area monitoring node to be measured is obtained as time series
The time series measurement data is inputted the UV254 prediction model based on Recognition with Recurrent Neural Network RNN, obtains institute by measurement data
State the UV254 prediction data at each measurement moment in the first duration of water area monitoring node to be measured；
Difference is calculated using the UV254 measurement data and UV254 prediction data, when obtaining each measurement in the first duration
The residual error at quarter generates confidence interval according to the distribution of the residual error；
If the residual error at measurement moment determines the measurement moment for normal point in the confidence interval in the first duration；
If the residual error at measurement moment is not in the confidence interval in the first duration, determine that the measurement moment is abnormal point；
For occurring in the second time window of all measurement moment computational representations in the second time window in the first duration, water quality is different
The cumulative probability of normal probability determines to have occurred in second time window if the cumulative probability is greater than predetermined probabilities threshold value
Water quality anomalous event.
The second aspect of the embodiment of the present invention provides a kind of device of water quality abnormality detection, comprising:
Prediction module, the UV254 for obtaining each measurement moment in the first duration of water area monitoring node to be measured measure number
According to as time series measurement data, it is pre- that the time series measurement data is inputted into the UV254 based on Recognition with Recurrent Neural Network RNN
Model is surveyed, the UV254 prediction data at each measurement moment in first duration of water area monitoring node to be measured is obtained；
Difference block, for calculating difference using the UV254 measurement data and UV254 prediction data, when obtaining first
The residual error at each measurement moment in long, generates confidence interval according to the distribution of the residual error；
First determination module, if the residual error for measuring the moment in the first duration in the confidence interval, determines to be somebody's turn to do
The measurement moment is normal point；If the residual error at measurement moment is not in the confidence interval in the first duration, the survey is determined
The amount moment is abnormal point；
Second determination module, for being directed in the first duration in the second time window when all measurement moment computational representations the second
Between the cumulative probability of water quality exception probability occurs in window, if the cumulative probability is greater than predetermined probabilities threshold value, determine described the
Water quality anomalous event has occurred in two time windows.
The third aspect of the embodiment of the present invention provides a kind of terminal device, including memory and processor, described to deposit
The computer program that can be run on the processor is stored in reservoir, when the processor executes the computer program,
The step of realizing method as described in relation to the first aspect.
The fourth aspect of the embodiment of the present invention provides a kind of computer readable storage medium, the computer-readable storage
Media storage has computer program, and the step of method as described in relation to the first aspect is realized when the computer program is executed by processor
Suddenly.
In the embodiment of the present invention, by being handled based on Recognition with Recurrent Neural Network (Recurrent neural network, RNN)
Time series data predicts UV254 data in waters to be measured, is calculated using UV254 measurement data and UV254 prediction data
Difference obtains residual error, determines abnormal point after generating confidence interval based on residual distribution, finally determines and sends water quality anomalous event
Time window.The embodiment of the present invention effectively analyzed and provide abnormality detection by the UV254 data based on one-dimensional time series
As a result, using the information in time series, in the case where lesser wrong report, the effective burst water pollution thing for detecting waters
Part, the biggish effect promoted to water area water-quality abnormality detection are preferable to river water quality anomalous identification effect.
Detailed description of the invention
It to describe the technical solutions in the embodiments of the present invention more clearly, below will be to embodiment or description of the prior art
Needed in attached drawing be briefly described, it should be apparent that, the accompanying drawings in the following description is only of the invention some
Embodiment for those of ordinary skill in the art without any creative labor, can also be according to these
Attached drawing obtains other attached drawings.
Fig. 1 is a kind of implementation flow chart of the method for water quality abnormality detection provided in an embodiment of the present invention；
Fig. 2 be the UV254 data in a kind of simulating, verifying provided in an embodiment of the present invention in time series true value with
The comparison schematic diagram of predicted value；
Fig. 3 is the residual error schematic diagram of UV254 data in time series in a kind of simulating, verifying provided in an embodiment of the present invention；
Fig. 4 is that predicted events position shows compared with actual event position in a kind of simulating, verifying provided in an embodiment of the present invention
It is intended to；
Fig. 5 is a kind of structural block diagram of the device of water quality abnormality detection provided in an embodiment of the present invention；
Fig. 6 is the schematic diagram of terminal device provided in an embodiment of the present invention.
Specific embodiment
In order to illustrate technical solutions according to the invention, the following is a description of specific embodiments.
It should be noted that in the absence of conflict, the features in the embodiments and the embodiments of the present application can phase
Mutually combination.The present invention will be described in detail below with reference to the accompanying drawings and embodiments.
In order to enable those skilled in the art to better understand the solution of the present invention, below in conjunction in the embodiment of the present invention
Attached drawing, technical solution in the embodiment of the present invention carry out clear, are fully described by, it is clear that described embodiment is only
The embodiment of a part of the invention, instead of all the embodiments.Based on the embodiments of the present invention, skill common for this field
For art personnel, without any creative labor, this hair is all should belong in every other embodiment obtained
The range of bright protection.
In being described below, for illustration and not for limitation, the tool of such as particular system structure, technology etc is proposed
Body details, to understand thoroughly the embodiment of the present invention.However, it will be clear to one skilled in the art that there is no these specific
The present invention also may be implemented in the other embodiments of details.In other situations, it omits to well-known system, device, electricity
The detailed description of road and method, in case unnecessary details interferes description of the invention.
It should be noted that involved in the term in specification of the invention, claims and attached drawing " first " or
The description of " second " etc. is only used for distinguishing similar object, is not understood to indicate or imply its relative importance or implicit
Indicate the quantity of indicated technical characteristic, that is to say, that these descriptions are not necessarily used for describing specific sequence or precedence.
Further, it will be understood that these descriptions are interchangeable under appropriate circumstances, to describe the embodiment of the present invention.
Fig. 1 shows the implementation process of the method for water quality abnormality detection provided in an embodiment of the present invention, this method process packet
Include step S101 to S104.This method is applicable to carry out waters the situation of water quality abnormality detection.This method is different by water quality
The device often detected executes, and the device of the water quality abnormality detection is configured at terminal device, can be implemented by software and/or hardware.
The specific implementation principle of each step is as follows.
S101, obtain in the first duration of water area monitoring node to be measured it is each measurement the moment UV254 measurement data as when
Between sequence measuring data, by the time series measurement data input the UV254 prediction model based on Recognition with Recurrent Neural Network RNN,
Obtain the UV254 prediction data at each measurement moment in first duration of water area monitoring node to be measured.
Wherein, waters to be measured is the object for needing to carry out water quality abnormality detection.Waters to be measured can with natural water area also with
For artificial waters, river, lake, the water supplying pipe and pond etc. of water factory and residential building can include but is not limited to.It retouches for convenience
It states, is illustrated in the following embodiments using river as waters to be measured.
Monitoring node is some test point in waters to be measured, and the embodiment of the present invention is not made to have to the position of the monitoring node
Body limits.
The UV254 data of the river water quality in the first duration of river monitoring node in time series are obtained as time series
Measurement data, at this point, getting the corresponding UV254 measurement data of each measurement moment in the first duration, it should be noted that each
A measurement moment and UV254 measurement data correspond.It should be noted that the first duration can be arranged according to demand, this hair
Bright embodiment is not especially limited this.
UV254 prediction model based on RNN carries out machine learning training using multiple groups sample data and obtains, for pre-
Survey UV254 data.Meeting concrete example illustrates the training process of RNN model in subsequent embodiment, refers to subsequent correlation
Description.
In order to make time series measurement data be suitable for RNN model, data are subjected to format conversion, are shown below:
Xn=[x (t), x (t-1) ..., x (t- (N-1))],
Wherein, x (t) is value of the UV254 data in time series in t moment, and N is the input number of the RNN model of construction
According to dimension, the expression formula for the one-dimensional matrix that above-mentioned data are constituted is as follows:
Optionally, the UV254 data of acquisition can also be pre-processed in other embodiments of the present invention, removes noise
Data, such as obvious unreasonable UV254 data.
S102 calculates difference using the UV254 measurement data and UV254 prediction data, obtains each in the first duration
The residual error for measuring the moment generates confidence interval according to the distribution of the residual error.
Wherein, the UV254 measurement data in river to be measured and UV254 prediction data are subjected to difference, obtain corresponding data it
Between residual error, that is to say, that each measurement moment corresponding residual error is obtained, then according to the distribution of residual error generation confidence area
Between.
Specifically, according to the distribution of the residual error generate confidence interval, comprising: by the first duration in first time window it is residual
Center μ of the mean value as normal distribution of difference, variances sigma of the standard deviation of residual error as normal distribution in first time window, according to
The center μ and variances sigma generate confidence interval.
It should be noted that first time window is empirical value, can be arranged according to demand, the embodiment of the present invention does not make this
It is specific to limit.Confidence interval is [μ-K σ, μ+K σ], wherein K value is any value in 0.4 to 0.7, can for 0.4,0.5,
0.6 or 0.7 etc..Rule of thumb, confidence interval is set in one embodiment of the invention as [+0.5 σ of μ -0.5 σ, μ].
S103, if the residual error at measurement moment determines that the measurement moment is positive in the confidence interval in the first duration
Chang Dian；If the residual error at measurement moment is not in the confidence interval in the first duration, determine that the measurement moment is exception
Point.
Wherein, when certain measurement the moment residual error in confidence interval, determine that the UV254 data are normal data, the measurement
Moment is normal point；When certain measurement the moment residual error not in confidence interval, determine that the UV254 data are abnormal data, the survey
The amount moment is abnormal point.
Optionally, in other embodiments of the present invention, after step 103, further includes: by the abnormal point in the first duration
Union is taken, the abnormal point set of the time series measurement data is obtained.
S104, for water occurs in the second time window of all measurement moment computational representations in the second time window in the first duration
The cumulative probability of matter exception probability determines hair in second time window if the cumulative probability is greater than predetermined probabilities threshold value
Water quality anomalous event is given birth to.
It wherein, will be abnormal at step 104 by carrying out sequence analysis to the data in the second time window in the first duration
Point is converted into unusual sequences analysis.
The cumulative probability at the second time window each measurement moment is calculated, cumulative probability occurs for characterizing in the second time window
The probability of water quality exception.When cumulative probability is more than given threshold, that is, determine that water quality anomalous event occurs for the second time window, also
To say, the second time window memory when abnormal between sequence.It should be noted that predetermined probabilities threshold value is empirical value, in the present invention
It being defined in embodiment not to this, predetermined probabilities threshold value can be any value in 0.8 to 0.9, can be 0.8 with value,
0.85 or 0.9.
Optionally, the calculation formula of the cumulative probability P (t+1) are as follows:
P (t+1)=ω0Yt+1+ω1Yt+....+ωnYt+1-n,
Wherein, preset weights ω0,...,ωnBe gradually reduced andWhen quarter t+1-i is abnormal point when measuring
Yt+1-i1 is taken, carves Y when t+1-i is normal point when measuringt+1-iTake 0.
The embodiment of the present invention effectively analyzed and provide abnormality detection by the UV254 data based on one-dimensional time series
As a result, using the information in time series, in the case where lesser wrong report, the effective burst water pollution thing for detecting waters
Part, the biggish effect promoted to water area water-quality abnormality detection are preferable to river water quality anomalous identification effect.
It will be illustrated below by concrete scheme of the embodiment to training RNN model.
On the basis of above-mentioned embodiment illustrated in fig. 1, the method also includes: use each measurement moment in the second duration
UV254 data as timed sample sequence measurement data, obtained by machine learning training based on Recognition with Recurrent Neural Network RNN's
UV254 prediction model.
Specifically, use the UV254 data at each measurement moment in the second duration as timed sample sequence measurement data,
The UV254 prediction model based on Recognition with Recurrent Neural Network RNN is obtained by machine learning training, comprising:
The UV254 data at each measurement moment in the second duration of sample water area monitoring node are obtained as timed sample sequence
The timed sample sequence measurement data is divided into training set sample data and test set sample data by measurement data, wherein instruction
White silk integrates the UV254 data in sample data all as the data of normal water matter；
The training set sample data and the test set sample data are input to RNN and be trained, and passes through grid
Method finds optimal hyper parameter, constructs the UV254 prediction model based on RNN.
Wherein, the second duration can be identical as the first duration, may also be distinct from that the first duration.
In order to make timed sample sequence measurement data be suitable for RNN model, data are subjected to format conversion, such as following formula institute
Show:
Wherein, x (t) is value of the UV254 data in time series in t moment, and N is the input number of the RNN model of construction
According to dimension, YnFor the output variable of RNN model, i.e. UV254 prediction data, the sample measurement at corresponding t+1 moment is above-mentioned
Data constitute two-dimensional matrix expression formula it is as follows:
Above-mentioned 2-D data is divided into training set sample data and test set sample data.Implement as the present invention one
Example, wherein 2/3 data are as training sample, 1/3 data are as test sample.
Above-mentioned training sample data are input to RNN model to be trained, determine the dimension of input data, the layer of network
Number, hidden layer node number, activation primitive, output layer and learning rate, and optimal hyper parameter is found using gridding method, obtain base
In the UV254 prediction model of the time series of RNN.
Optionally, the number of plies of network structure is determined according to the dimension of the complexity of UV254 data and input data.Hidden
Select linear function as activation primitive between hiding layer and output layer
Optionally, the UV254 prediction model based on RNN is to remember (LongShort-Term based on shot and long term
Memory, LSTM) network UV254 prediction model.
Wherein, the number of plies that LSTM network structure is determined according to the dimension of the complexity of UV254 data and input data,
Select linear function as activation primitive between hidden layer and output layer.Since LSTM network can preferably utilize time series
On data feature, can further promote the accuracy of water quality abnormality detection.
In the following, abnormal water detection method based on the embodiment of the present invention, according to certain node in certain city river
Historical Monitoring UV254 data be simulating, verifying that example carries out algorithm, obtain the historical data in certain time, in total 1500
The data of group or so；And by the way that 2 organic pollutant intrusion events, 20 sampled points of event duration, as test are manually set
Data.The UV254 prediction model based on LSTM network is used in this simulating, verifying.Fig. 2 show the time in simulating, verifying
The comparison schematic diagram of the predicted value of the measured value (i.e. true value) and model of UV254 data in sequence.Fig. 3 show time sequence
The residual error that measured value (i.e. true value) and predicted value difference obtain on column, " confidence interval 1 " refers to the upper of confidence interval in figure
Boundary, " confidence interval 2 " refer to the lower boundary of confidence interval, and the K value of confidence interval is 0.5 at this time.Fig. 4 show reality
Time location is compared with predicted events position, and event is water quality anomalous event, and predetermined probabilities threshold value is set as 0.85 at this time.It is logical
Crossing simulating, verifying can determine that water quality anomalous event frequency is that twice, the position of predicted events is also quite accurate, detection effect
Fruit is fine.
It should be understood that the size of the serial number of each step is not meant that the order of the execution order in above-described embodiment, each process
Execution sequence should be determined by its function and internal logic, the implementation process without coping with the embodiment of the present invention constitutes any limit
It is fixed.
Corresponding to the method for water quality abnormality detection described in foregoing embodiments, Fig. 5 shows provided in an embodiment of the present invention
The structural block diagram of the device of water quality abnormality detection, for ease of description, only parts related to embodiments of the present invention are shown.
Referring to Fig. 5, the device of the water quality abnormality detection includes:
Prediction module 51, for obtaining the UV254 measurement at each measurement moment in the first duration of water area monitoring node to be measured
Data input the UV254 based on Recognition with Recurrent Neural Network RNN as time series measurement data, by the time series measurement data
Prediction model obtains the UV254 prediction data at each measurement moment in first duration of water area monitoring node to be measured；
Difference block 52 obtains first for calculating difference using the UV254 measurement data and UV254 prediction data
The residual error at each measurement moment in duration, generates confidence interval according to the distribution of the residual error；
First determination module 53, if the residual error for measuring the moment in the first duration determines in the confidence interval
The measurement moment is normal point；If the residual error at measurement moment is not in the confidence interval in the first duration, described in judgement
The measurement moment is abnormal point；
Second determination module 54, for for all measurement moment computational representations second in the second time window in the first duration
The cumulative probability of water quality exception probability occurs in time window, if the cumulative probability is greater than predetermined probabilities threshold value, determine described in
Water quality anomalous event has occurred in second time window.
It is optionally, described that confidence interval is generated according to the distribution of the residual error, comprising:
The residual error for obtaining each measurement moment in the first duration, by the mean value of residual error is made in first time window in the first duration
For the center μ of normal distribution, variances sigma of the standard deviation of residual error as normal distribution in first time window, according to the center μ and
Variances sigma generates confidence interval.
Optionally, the calculation formula of the cumulative probability P (t+1) are as follows:
P (t+1)=ω0Yt+1+ω1Yt+....+ωnYt+1-n,
Wherein, preset weights ω0,...,ωnBe gradually reduced andWhen quarter t+1-i is abnormal point when measuring
Yt+1-i1 is taken, carves Y when t+1-i is normal point when measuringt+1-iTake 0.
Optionally, further include training module, for use in the second duration it is each measurement the moment UV254 data as when
Between sequence samples measurement data, by machine learning training obtain the UV254 prediction model based on Recognition with Recurrent Neural Network RNN.
Optionally, training module is specifically used for:
The UV254 data at each measurement moment in the second duration of sample water area monitoring node are obtained as timed sample sequence
The timed sample sequence measurement data is divided into training set sample data and test set sample data by measurement data, wherein instruction
White silk integrates the UV254 data in sample data all as the data of normal water matter；
The training set sample data and the test set sample data are input to RNN and be trained, and passes through grid
Method finds optimal hyper parameter, constructs the UV254 prediction model based on RNN.
Fig. 6 is the schematic diagram for the terminal device that one embodiment of the invention provides.As shown in fig. 6, the terminal of the embodiment is set
Standby 6 include: processor 60, memory 61 and are stored in the meter that can be run in the memory 61 and on the processor 60
Calculation machine program 62, such as the program of water quality abnormality detection.The processor 60 is realized above-mentioned when executing the computer program 62
Step in the embodiment of the method for water quality abnormality detection, such as step S101 to S104 shown in FIG. 1.Alternatively, the processor
The function of each module/unit in above-mentioned each Installation practice, such as mould shown in Fig. 5 are realized when the 60 execution computer program 62
The function of block 51 to 54.
Illustratively, the computer program 62 can be divided into one or more module/units, it is one or
Multiple module/units are stored in the memory 61, and are executed by the processor 60, to complete the present invention.Described one
A or multiple module/units can be the series of computation machine program instruction section that can complete specific function, which is used for
Implementation procedure of the computer program 62 in the terminal device 6 is described.
The terminal device 6 can be smart phone, computer, plate, server etc..The terminal device 6 may include, but
It is not limited only to, processor 60, memory 61.It will be understood by those skilled in the art that Fig. 6 is only the example of terminal device 6, and
Do not constitute the restriction to terminal device 6, may include than illustrating more or fewer components, perhaps combine certain components or
Different components, such as the terminal device can also include input-output equipment, network access equipment, bus etc..
Alleged processor 60 can be central processing unit (Central Processing Unit, CPU), can also be
Other general processors, digital signal processor (Digital Signal Processor, DSP), specific integrated circuit
(Application Specific Integrated Circuit, ASIC), field programmable gate array (Field-
Programmable Gate Array, FPGA) either other programmable logic device, discrete gate or transistor logic,
Discrete hardware components etc..General processor can be microprocessor or the processor is also possible to any conventional processor
Deng.
The memory 61 can be the internal storage unit of the terminal device 6, such as the hard disk or interior of terminal device 6
It deposits.The memory 61 is also possible to the External memory equipment of the terminal device 6, such as be equipped on the terminal device 6
Plug-in type hard disk, intelligent memory card (Smart Media Card, SMC), secure digital (Secure Digital, SD) card dodge
Deposit card (Flash Card) etc..Further, the memory 61 can also both include the storage inside list of the terminal device 6
Member also includes External memory equipment.The memory 61 is for storing needed for the computer program and the terminal device
Other programs and data.The memory 61 can be also used for temporarily storing the data that has exported or will export.
It is apparent to those skilled in the art that for convenience of description and succinctly, only with above-mentioned each function
Can unit, module division progress for example, in practical application, can according to need and by above-mentioned function distribution by different
Functional unit, module are completed, i.e., the internal structure of described device is divided into different functional unit or module, more than completing
The all or part of function of description.Each functional unit in embodiment, module can integrate in one processing unit, can also
To be that each unit physically exists alone, can also be integrated in one unit with two or more units, it is above-mentioned integrated
Unit both can take the form of hardware realization, can also realize in the form of software functional units.In addition, each function list
Member, the specific name of module are also only for convenience of distinguishing each other, the protection scope being not intended to limit this application.Above system
The specific work process of middle unit, module, can refer to corresponding processes in the foregoing method embodiment, and details are not described herein.
In the above-described embodiments, it all emphasizes particularly on different fields to the description of each embodiment, is not described in detail or remembers in some embodiment
The part of load may refer to the associated description of other embodiments.
The unit as illustrated by the separation member may or may not be physically separated, aobvious as unit
The component shown may or may not be physical unit, it can and it is in one place, or may be distributed over multiple
In network unit.It can select some or all of unit therein according to the actual needs to realize the mesh of this embodiment scheme
's.
It, can also be in addition, the functional units in various embodiments of the present invention may be integrated into one processing unit
It is that each unit physically exists alone, can also be integrated in one unit with two or more units.Above-mentioned integrated list
Member both can take the form of hardware realization, can also realize in the form of software functional units.
If the integrated module/unit be realized in the form of SFU software functional unit and as independent product sale or
In use, can store in a computer readable storage medium.Based on this understanding, the present invention realizes above-mentioned implementation
All or part of the process in example method, can also instruct relevant hardware to complete, the meter by computer program
Calculation machine program can be stored in a computer readable storage medium, the computer program when being executed by processor, it can be achieved that on
The step of stating each embodiment of the method.
Embodiment described above is merely illustrative of the technical solution of the present invention, rather than its limitations；Although referring to aforementioned reality
Applying example, invention is explained in detail, those skilled in the art should understand that: it still can be to aforementioned each
Technical solution documented by embodiment is modified or equivalent replacement of some of the technical features；And these are modified
Or replacement, the spirit and scope for technical solution of various embodiments of the present invention that it does not separate the essence of the corresponding technical solution should all
It is included within protection scope of the present invention.
Claims (10)
1. a kind of method of water quality abnormality detection characterized by comprising
The UV254 measurement data for obtaining each measurement moment in the first duration of water area monitoring node to be measured is measured as time series
The time series measurement data is inputted the UV254 prediction model based on Recognition with Recurrent Neural Network RNN by data, obtain it is described to
Survey the UV254 prediction data at each measurement moment in the first duration of water area monitoring node；
Difference is calculated using the UV254 measurement data and UV254 prediction data, obtains each measurement moment in the first duration
Residual error generates confidence interval according to the distribution of the residual error；
If the residual error at measurement moment determines the measurement moment for normal point in the confidence interval in the first duration；If the
When the residual error at measurement moment is not in the confidence interval in one duration, then determine that the measurement moment is abnormal point；
For occurring in the second time window of all measurement moment computational representations in the second time window in the first duration, water quality is extremely general
The cumulative probability of rate if the cumulative probability is greater than predetermined probabilities threshold value determines that water quality has occurred in second time window
Anomalous event.
2. the method as described in claim 1, which is characterized in that described to generate confidence interval, packet according to the distribution of the residual error
It includes:
Obtain in the first duration it is each measurement the moment residual error, using in the first duration in first time window the mean value of residual error as just
The center μ of state distribution, variances sigma of the standard deviation of residual error as normal distribution in first time window, according to the center μ and variance
σ generates confidence interval.
3. method according to claim 1 or 2, which is characterized in that the calculation formula of the cumulative probability P (t+1) are as follows:
P (t+1)=ω0Yt+1+ω1Yt+....+ωnYt+1-n,
Wherein, preset weights ω0,...,ωnBe gradually reduced andY when t+1-i is abnormal point is carved when measuringt+1-iIt takes
1, Y when t+1-i is normal point is carved when measuringt+1-iTake 0.
4. method according to claim 1 or 2, which is characterized in that further include: use the measurement moment each in the second duration
UV254 data are obtained by machine learning training based on Recognition with Recurrent Neural Network RNN's as timed sample sequence measurement data
UV254 prediction model.
5. method as claimed in claim 4, which is characterized in that the UV254 using the measurement moment each in the second duration
It is pre- that data as timed sample sequence measurement data, by machine learning training obtain the UV254 based on Recognition with Recurrent Neural Network RNN
Survey model, comprising:
The UV254 data for obtaining each measurement moment in the second duration of sample water area monitoring node are measured as timed sample sequence
The timed sample sequence measurement data is divided into training set sample data and test set sample data, wherein training set by data
UV254 data in sample data are all the data of normal water matter；
The training set sample data and the test set sample data are input to RNN and be trained, and is sought by gridding method
Optimal hyper parameter is looked for, the UV254 prediction model based on RNN is constructed.
6. such as the described in any item methods of claim 1,2 or 5, which is characterized in that the UV254 prediction model based on RNN
For the UV254 prediction model for remembering LSTM network based on shot and long term.
7. method as claimed in claim 6, which is characterized in that the parameter of the RNN includes: the dimension of input data, network
The number of plies, hidden layer node number, activation primitive and learning rate.
8. a kind of device of water quality abnormality detection characterized by comprising
Prediction module, the UV254 measurement data for obtaining each measurement moment in the first duration of water area monitoring node to be measured are made
For time series measurement data, the time series measurement data is inputted into the UV254 based on Recognition with Recurrent Neural Network RNN and predicts mould
Type obtains the UV254 prediction data at each measurement moment in first duration of water area monitoring node to be measured；
Difference block obtains in the first duration for calculating difference using the UV254 measurement data and UV254 prediction data
The residual error at each measurement moment generates confidence interval according to the distribution of the residual error；
First determination module, if the residual error for measuring the moment in the first duration determines the measurement in the confidence interval
Moment is normal point；If the residual error at measurement moment is not in the confidence interval in the first duration, when determining the measurement
Carving is abnormal point；
Second determination module, for for the second time window of all measurement moment computational representations in the second time window in the first duration
The interior cumulative probability that water quality exception probability occurs, if the cumulative probability is greater than predetermined probabilities threshold value, when determining described second
Between water quality anomalous event has occurred in window.
9. a kind of terminal device, including memory and processor, it is stored with and can transports on the processor in the memory
Capable computer program, which is characterized in that when the processor executes the computer program, realize such as claim 1 to 7 times
The step of one the method.
10. a kind of computer readable storage medium, the computer-readable recording medium storage has computer program, and feature exists
In when the computer program is executed by processor the step of any one of such as claim 1 to 7 of realization the method.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910497097.4A CN110231447A (en) | 2019-06-10 | 2019-06-10 | The method, apparatus and terminal device of water quality abnormality detection |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910497097.4A CN110231447A (en) | 2019-06-10 | 2019-06-10 | The method, apparatus and terminal device of water quality abnormality detection |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110231447A true CN110231447A (en) | 2019-09-13 |
Family
ID=67859599
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910497097.4A Pending CN110231447A (en) | 2019-06-10 | 2019-06-10 | The method, apparatus and terminal device of water quality abnormality detection |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110231447A (en) |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110835128A (en) * | 2019-11-26 | 2020-02-25 | 倪世章 | Water purifier and method for purifying water quality in real time based on water quality |
| CN110888788A (en) * | 2019-10-16 | 2020-03-17 | 平安科技（深圳）有限公司 | Anomaly detection method and device, computer equipment and storage medium |
| CN110969556A (en) * | 2019-09-30 | 2020-04-07 | 上海仪电（集团）有限公司中央研究院 | Method and device for detecting river water quality abnormity by machine learning multi-dimension multi-model fusion |
| CN111191855A (en) * | 2020-01-13 | 2020-05-22 | 大连理工大学 | Water quality abnormal event identification and early warning method based on pipe network multi-element water quality time sequence data |
| CN111275307A (en) * | 2020-01-16 | 2020-06-12 | 生态环境部华南环境科学研究所 | Quality control method for high-frequency continuous observation data of automatic online water quality station |
| CN113535486A (en) * | 2020-04-14 | 2021-10-22 | 北京沃东天骏信息技术有限公司 | Method and device for testing performance of machine |
| CN114022052A (en) * | 2022-01-04 | 2022-02-08 | 北京英视睿达科技股份有限公司 | Water quality abnormity monitoring method and device, storage medium and computer equipment |
| CN116304913A (en) * | 2023-04-07 | 2023-06-23 | 中国长江三峡集团有限公司 | Water quality state monitoring method and device based on Bayesian model and electronic equipment |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2013088981A (en) * | 2011-10-17 | 2013-05-13 | Chugoku Electric Power Co Inc:The | Time prediction device, time prediction method and program |
| CN104034684A (en) * | 2014-06-05 | 2014-09-10 | 北京金达清创环境科技有限公司 | Water quality multi-index detection method on basis of ultraviolet-visible absorption spectrum |
| CN106489068A (en) * | 2014-03-31 | 2017-03-08 | 株式会社Ihi | Measured value analytical equipment and measured value analysis method |
| CN108089962A (en) * | 2017-11-13 | 2018-05-29 | 北京奇艺世纪科技有限公司 | A kind of method for detecting abnormality, device and electronic equipment |
| CN108167653A (en) * | 2017-12-23 | 2018-06-15 | 浙江大学 | Public supply mains booster detection method based on dynamic neural network prediction |
| CN108760642A (en) * | 2018-08-22 | 2018-11-06 | 成都益清源科技有限公司 | The real-time Water Test Kits of full spectrum |
| CN108900546A (en) * | 2018-08-13 | 2018-11-27 | 杭州安恒信息技术股份有限公司 | The method and apparatus of time series Network anomaly detection based on LSTM |
| CN109495327A (en) * | 2018-12-28 | 2019-03-19 | 西安交通大学 | A kind of User Activity abnormality detection and method for predicting based on big data analysis |
| CN109641602A (en) * | 2017-07-14 | 2019-04-16 | 株式会社东芝 | Abnormality detecting apparatus, method for detecting abnormality and non-transitory computer-readable medium |
| US20190147343A1 (en) * | 2017-11-15 | 2019-05-16 | International Business Machines Corporation | Unsupervised anomaly detection using generative adversarial networks |
-
2019
- 2019-06-10 CN CN201910497097.4A patent/CN110231447A/en active Pending
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2013088981A (en) * | 2011-10-17 | 2013-05-13 | Chugoku Electric Power Co Inc:The | Time prediction device, time prediction method and program |
| CN106489068A (en) * | 2014-03-31 | 2017-03-08 | 株式会社Ihi | Measured value analytical equipment and measured value analysis method |
| CN104034684A (en) * | 2014-06-05 | 2014-09-10 | 北京金达清创环境科技有限公司 | Water quality multi-index detection method on basis of ultraviolet-visible absorption spectrum |
| CN109641602A (en) * | 2017-07-14 | 2019-04-16 | 株式会社东芝 | Abnormality detecting apparatus, method for detecting abnormality and non-transitory computer-readable medium |
| CN108089962A (en) * | 2017-11-13 | 2018-05-29 | 北京奇艺世纪科技有限公司 | A kind of method for detecting abnormality, device and electronic equipment |
| US20190147343A1 (en) * | 2017-11-15 | 2019-05-16 | International Business Machines Corporation | Unsupervised anomaly detection using generative adversarial networks |
| CN108167653A (en) * | 2017-12-23 | 2018-06-15 | 浙江大学 | Public supply mains booster detection method based on dynamic neural network prediction |
| CN108900546A (en) * | 2018-08-13 | 2018-11-27 | 杭州安恒信息技术股份有限公司 | The method and apparatus of time series Network anomaly detection based on LSTM |
| CN108760642A (en) * | 2018-08-22 | 2018-11-06 | 成都益清源科技有限公司 | The real-time Water Test Kits of full spectrum |
| CN109495327A (en) * | 2018-12-28 | 2019-03-19 | 西安交通大学 | A kind of User Activity abnormality detection and method for predicting based on big data analysis |
Non-Patent Citations (1)
| Title |
|---|
| 张清周: "基于水力模型的供水管网压力管理与漏损区域识别研究", 《中国博士学位论文全文数据库 工程科技Ⅱ辑》 * |
Cited By (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110969556A (en) * | 2019-09-30 | 2020-04-07 | 上海仪电（集团）有限公司中央研究院 | Method and device for detecting river water quality abnormity by machine learning multi-dimension multi-model fusion |
| CN110969556B (en) * | 2019-09-30 | 2023-11-21 | 上海仪电（集团）有限公司中央研究院 | Machine learning multidimensional multi-model fusion river channel water quality anomaly detection method and device |
| CN110888788A (en) * | 2019-10-16 | 2020-03-17 | 平安科技（深圳）有限公司 | Anomaly detection method and device, computer equipment and storage medium |
| CN110888788B (en) * | 2019-10-16 | 2024-09-10 | 平安科技（深圳）有限公司 | Abnormality detection method, abnormality detection device, computer device, and storage medium |
| CN110835128B (en) * | 2019-11-26 | 2021-12-07 | 陈文印 | Water purifier and method for purifying water quality in real time based on water quality |
| CN110835128A (en) * | 2019-11-26 | 2020-02-25 | 倪世章 | Water purifier and method for purifying water quality in real time based on water quality |
| CN111191855A (en) * | 2020-01-13 | 2020-05-22 | 大连理工大学 | Water quality abnormal event identification and early warning method based on pipe network multi-element water quality time sequence data |
| CN111191855B (en) * | 2020-01-13 | 2022-09-20 | 大连理工大学 | Water quality abnormal event identification and early warning method based on pipe network multi-element water quality time sequence data |
| CN111275307B (en) * | 2020-01-16 | 2023-09-05 | 生态环境部华南环境科学研究所 | Quality control method for high-frequency continuous observation data of automatic online station for water quality |
| CN111275307A (en) * | 2020-01-16 | 2020-06-12 | 生态环境部华南环境科学研究所 | Quality control method for high-frequency continuous observation data of automatic online water quality station |
| CN113535486A (en) * | 2020-04-14 | 2021-10-22 | 北京沃东天骏信息技术有限公司 | Method and device for testing performance of machine |
| CN114022052A (en) * | 2022-01-04 | 2022-02-08 | 北京英视睿达科技股份有限公司 | Water quality abnormity monitoring method and device, storage medium and computer equipment |
| CN116304913A (en) * | 2023-04-07 | 2023-06-23 | 中国长江三峡集团有限公司 | Water quality state monitoring method and device based on Bayesian model and electronic equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110231447A (en) | The method, apparatus and terminal device of water quality abnormality detection | |
| Ciancetta et al. | A new convolutional neural network-based system for NILM applications | |
| Fedra et al. | Uncertainty and arbitrariness in ecosystems modelling: a lake modelling example | |
| CN110247910A (en) | A kind of detection method of abnormal flow, system and associated component | |
| CN109345553A (en) | A kind of palm and its critical point detection method, apparatus and terminal device | |
| CN105373830A (en) | Prediction method and system for error back propagation neural network and server | |
| Vitter et al. | A non-intrusive approach for classifying residential water events using coincident electricity data | |
| Sun et al. | Contamination source identification in water distribution networks using convolutional neural network | |
| CN116341901A (en) | Integrated evaluation method for landslide surface domain-monomer hazard early warning | |
| Medeiros et al. | Ranking species based on sensitivity to perturbations under non‐equilibrium community dynamics | |
| CN117115637A (en) | Water quality monitoring and early warning method and system based on big data technology | |
| CN115545334A (en) | Land use type prediction method, land use type prediction device, electronic device, and storage medium | |
| CN114742183A (en) | Water pollution early warning method, system, equipment and storage medium based on deep learning | |
| Parviz et al. | Development of precipitation forecast model based on artificial intelligence and subseasonal clustering | |
| Dawson et al. | Sensitivity analysis for comparison, validation and physical legitimacy of neural network-based hydrological models | |
| Wiryasaputra et al. | An IoT Real-Time Potable Water Quality Monitoring and Prediction Model Based on Cloud Computing Architecture | |
| Tarasov et al. | Review and possible development direction of the methods for modeling of soil pollutants spatial distribution | |
| Ju et al. | Hydrologic simulations with artificial neural networks | |
| CN116432839A (en) | Backfill cost prediction method, device and equipment based on BP neural network | |
| CN116484219A (en) | Water supply network water quality abnormal pollution source identification method based on gate control graph neural network | |
| CN109726824A (en) | The transfer learning method and terminal device of training pattern | |
| CN114565286A (en) | Intelligent analysis method and device for industrial wastewater discharge | |
| CN113011657A (en) | Method and device for predicting typhoon water level, electronic equipment and storage medium | |
| CN106772306B (en) | A kind of detection method and server of object | |
| Gan et al. | A potential method for determining nonlinearity in wind data |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190913 |
|
| RJ01 | Rejection of invention patent application after publication |