CN102637019B - Intelligent integrated fault diagnosis method and device in industrial production process - Google Patents
Intelligent integrated fault diagnosis method and device in industrial production process Download PDFInfo
- Publication number
- CN102637019B CN102637019B CN201110035484.XA CN201110035484A CN102637019B CN 102637019 B CN102637019 B CN 102637019B CN 201110035484 A CN201110035484 A CN 201110035484A CN 102637019 B CN102637019 B CN 102637019B
- Authority
- CN
- China
- Prior art keywords
- fault
- analysis
- diagnosis
- fault diagnosis
- fmea
- 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.)
- Expired - Fee Related
Links
Images
Classifications
-
- 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
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/02—Total factory control, e.g. smart factories, flexible manufacturing systems [FMS] or integrated manufacturing systems [IMS]
Landscapes
- Test And Diagnosis Of Digital Computers (AREA)
- Testing And Monitoring For Control Systems (AREA)
Abstract
The invention relates to an intelligent integrated fault diagnosis method in an industrial production process. The intelligent integrated fault diagnosis method is characterized by comprising the following steps of: acquiring data in the industrial production process; analyzing and processing object characteristics according to an acquired signal; combining expert knowledge according to an intelligent integration method to carry out blast-furnace fault diagnosis analysis so as to identify a fault and find out a reason of the fault, carrying out fault exact location and diagnosis policy, and effectively regulating a production process so that the industrial production process can regularly carry out, wherein the intelligent integration method comprises the following steps of: establishing a Bayesian network model; comprehensively analyzing and processing FTA (full type approval) and FMEA (failure mode and effect analysis) models; and carrying out nerve net expert system fault diagnosis analysis and process. Simultaneously, the invention further relates to an intelligent integrated fault diagnosis device in the industrial production process, and the device is used for realizing the fault diagnosis method. According to the intelligent integrated fault diagnosis method in the industrial production process, disclosed by the invention, various information is fused, ratiocination is carried out under a complex situation, comprehensive diagnosis can be effectively carried out on the fault of the industrial production process, the integration, intelligence, accuracy and effectiveness of the fault diagnosis system are improved, and the production process is ensured to be performed smoothly.
Description
Technical field
The present invention relates to fault diagnosis technology in observation and control technology, relate in particular to the intelligent integrated diagnosis method in fault diagnosis, and relate more specifically to the industrial processes integrated method for diagnosing faults of intelligence and implement device.
Background technology
Along with modern industry and scientific and technical development, because product quality is improved, output improves, save the energy, antipollution needs, for continuously, large batch of modern production process is carried out fault diagnosis in the urgent need to setting up supervisory system to industries process control system, detect in real time phylogenetic fault, and to failure cause, failure-frequency and fault harm degree are analyzed, judgement, reach a conclusion, take necessary measure, the generation of incidents is against any misfortune, ensure the stable operation of producing, thereby conscientiously improve the economic benefit of industrial enterprise.
Existing method for diagnosing faults mainly contains: rule-based diagnosis, the diagnosis based on example, fuzzy diagnosis etc., various diagnostic methods obtain at diagnostic knowledge, the reliability of diagnosis, and in interpretability, have feature separately.
Rule-based diagnostic method is to set up by the accumulation of expert diagnosis experience.These experiences are described by rule format, and sign and potential fault are connected to this diagnostic method.Patent documentation: a kind of method (CN100393048C) of network fault diagnosis rule base, a kind of method for diagnosing faults and system (CN101848477A) and air-conditioning system method for diagnosing faults (CN1967077A) based on sort-type rule set up, by setting up network fault diagnosis rule base, knowledge base is carried out fault diagnosis.The problems such as the method exists knowledge acquisition difficulty, and between knowledge, upper and lower literary composition is responsive, and uncertain reasoning and adaptive ability are poor.Aspect diagnosis, the problem of existence is strong to the dependence of production system, and each new production system all needs one group of new regulation, need to be for a long time and accumulate these rules, and knowledge has become the encapsulation of experience.Before fault diagnosis, must there is considerable empirical rule, otherwise cannot diagnose.If diagnostic experiences is few, not only there will be news, also can fail to pinpoint a disease in diagnosis phenomenon.
Diagnosis based on example is that a kind of experience example in the past that uses instructs the method that solves new problem, and the key of this diagnostic method, is how to set up an effective example Indexing Mechanism and example organization mode.Patent documentation: a kind of based on PDA fault diagnosis system and method (CN101387582A), mainly for flight system, adopt typical case method to carry out fault analysis, the diagnosis example that the method can be collected is limited, can not cover all solution spaces, during search, also may miss optimum solution, when there is abnormal sign, owing to can not find optimum matching, may cause mistaken diagnosis or fail to pinpoint a disease in diagnosis, producing serious consequence.In addition, the consistency maintenance difficulty between fault diagnosis example, rewrites new example and needs more knowledge.A little less than the fault diagnosis system antijamming capability of setting up like this, security is difficult to ensure.
Diagnostic method based on fuzzy theory, because fuzzy language variable approaches natural language, the expression of knowledge is readable strong, and fuzzy reasoning logic is rigorous, and similar human thinking's process is easy to explain.But fuzzy diagnosis knowledge acquisition is difficult, especially the fuzzy relation of fault and sign is more difficult determines, and the diagnosis capability of system dependence fuzzy knowledge base, and learning ability is poor, mistaken diagnosis easily occurs or fail to pinpoint a disease in diagnosis.
Because current various method for diagnosing faults exist limitation separately, fault diagnosis for this complex object of industrial processes, as the knowledge representation method with single, sometimes be difficult to the intactly fault diagnosis ken of indicated object, integrated Several Knowledge Representations is the fault diagnosis field knowledge of indicated object better.The feature of the comprehensive various diagnostic methods of integrated diagnosis method energy based on Bayesian network model, FTA fault tree analysis, FMEA Failure Mode Effective Analysis model, neural network model, expert system, overcome the limitation of each diagnostic method, thereby the intelligent and diagnosis efficiency that improves diagnostic system, the fault diagnosis system of integrated-type can also be processed truth maintenance, conclusion explanation, machine learning etc. effectively.The integrated fault diagnosis technology of intelligence is an important development direction of fault diagnosis field.
At present the device of fault diagnosis adopts conventional instrument and equipment and traditional industrial control unit (ICU) more, and the conventional simulating signal diagnostic device functional structures of these tradition are single, and reliability, interchangeability are poor, difficult in maintenance, has affected the effect of fault diagnosis.The fault diagnosis that the present invention adopts totally digitilized FCS field bus control system to carry out industrial processes, because this system is-individual distributed network control system, function synthesized, reliability is high, interchangeability good, antijamming capability is strong, maintenance is easy, install and use the features such as expense is low, can effectively improve fault diagnosis efficiency, system reaches good diagnosis control level.
The method has important using value, and the phenomenon that it can more effectively reduce fault misdescription or fail to report takes measures correspondingly to adjust in time, ensures that industrial processes move safely and effectively.
Summary of the invention
The technical problem to be solved in the present invention is, for features such as the complicacy of industrial processes, uncertainties, and the limitation of existing diagnostic techniques and method existence, by accurate, reliable mode, carry out industrial process fault diagnosis, for providing effectively, normal industrial processes ensure.
For achieving the above object, the invention provides a kind of industrial process fault diagnosis method, it is characterized in that comprising the following steps:
The work such as the detection of production run fault diagnosis data, signals collecting;
According to the signal collecting, carry out analysis and the processing of characteristics of objects;
According to intelligent integrated approach, according to characteristics of objects, carry out production run Analysis on Fault Diagnosis, fault is identified, and trouble-shooting reason, carries out fault and accurately locates, and diagnose decision-making, effectively carry out system adjusting, thereby industrial processes can be carried out smoothly.
Wherein intelligent integrated diagnosis method comprises following steps:
The foundation of Bayesian network model;
Comprehensive analysis and the processing of FTA (fault tree analysis) and FMEA (Failure Mode Effective Analysis) model;
Neural network model Analysis on Fault Diagnosis and processing;
Expert system Analysis on Fault Diagnosis and processing.
The integrated fault diagnosis technology scheme of intelligence: whole network diagnostic systems consists of Bayesian network model, FTA and FMEA failure effect analysis (FEA) model and neural network model, in order to complete foundation, fault analysis and the reasoning from logic process of model.Here introducing Bayesian network model is in order to solve the uncertain factor of industrial processes, improves the accuracy of industrial process fault diagnosis; Adopt FTA and FMEA technology, can help system determine in early days in fault diagnosis the module that causes software failure, dwindle the scope of localization of fault.In addition, Bayesian network technology has very large similarity with FTA and FMEA in inference mechanism and system state, simultaneously can also improve their descriptive power, the probabilistic relation between, random occurrence uncertain by expressing obtains how useful conclusion, for fault diagnosis and location provides foundation.
Here FMEA Failure Mode Effective Analysis method be each module in analytic system, a kind of inductive method likely affecting setting up all issuable fault modes and system is caused; FTA fault tree analysis is the logical relation for showing which module of system has fault, external event or their combination to cause system to break down.Use separately FTA and FMEA all respectively to have drawback, only utilize FMEA can strengthen workload, and easily omit fault mode and fault effects, only utilize FTA easily to omit again the bottom event that causes top event to occur.Therefore, be the drawback of avoiding using separately FTA and FMEA to bring, FTA and FMEA are combined with Bayesian network technology respectively, turn to bayesian network structure model and carry out fault diagnosis, can improve fault diagnosis efficiency.First diagnostic model is analyzed phenomenon of the failure by the FTA based on Bayesian network and FMEA Comprehensive Analysis Technique, fault Primary Location is arrived to a certain module, then use neural network expert system diagnostic techniques to realize fault and further accurately locate, complete failure diagnostic process.
We adopt neural network by elementary knowledge being learnt the bottom fault of processing procedure, target level knowledge and the high-rise diagnosing malfunction of intergrade knowledge to process that expert system stores according to knowledge base.
Whole integrated system makes full use of model internal fault diagnostic method advantage separately and carries out complementation, thereby effectively overcome the deficiency that they exist separately, has good diagnosis efficiency and reliability.
The present invention adopts a kind of device of industrial process fault diagnosis, and this device field bus control system FCS is a network control system, and for realizing intelligent integrated method for diagnosing faults, it comprises:
FD fieldbus instrument and equipment, for input, output, computing, the control of fault diagnosis data-signal with communicate by letter, and provide functional block, to form control loop in bus at the scene;
FBI field-bus interface, under connect FNET fieldbus networks, on connect SNET monitor network, thereby realize interconnecting of FNET fieldbus networks and SNET monitor network;
FNET fieldbus networks, under connect FD fieldbus instrument and equipment, on connect FBI field-bus interface, carry out the mutual exchange of on-site signal and control signal;
SNET monitor network, for connecting FBI field-bus interface, OS operator's computing machine, ES slip-stick artist's controller, SCS fault diagnosis supervisory control comuter and CG computing machine gateway, carries out signal transmission;
OS operator's computing machine, monitors, operates and manage production run for site technique operator, carries out man-machine conversation;
ES slip-stick artist's controller, carries out system generation for field engineer to FCS and fault diagnosis is safeguarded, the establishment of control engineering Shi Jinhang control loop configuration programming and special applications software is provided;
SCS fault diagnosis supervisory control comuter, for foundation and the updating maintenance of foundation, database and the knowledge base of fault diagnosis model, production run is carried out to fault diagnosis, forecast and analysis, guarantees safety in production;
CG computing machine gateway, for connecting monitor network and production management network, realizes the intercommunication mutually between them.
When trouble-shooter field bus control system FCS moves, first by FD fieldbus instrument and equipment, gather the necessary diagnostic message of fault diagnosis and failure symptom, by FBI field-bus interface and SNET monitor network, information is uploaded to fault diagnosis supervisory control comuter, by it according to diagnostic knowledge, utilize intelligent integrated model for fault diagnosis to carry out reasoning, according to the bottom fault knowledge being stored on network weight, complete to detecting the fault analysis of data, with the consequent output of failure symptom, inference machine receives after the bottom fault diagnosis result of network diagnostic systems, in conjunction with the target level knowledge in knowledge base, strategy carries out reasoning diagnosis by inference, obtain the prediction conclusion of high-rise fault diagnosis, if prediction fault is identified, by elementary knowledge, instruct inference machine from prediction conclusion, utilize target level knowledge by inference strategy carry out reasoning diagnosis, if the failure symptom that reasoning draws is consistent with the failure symptom of network diagnosis model, high-rise fault is identified and is sent to interpretation process system, interpretation process system points out according to elementary knowledge the position that reason that fault may occur and fault occur, and provide the correlation method of handling failure, by slip-stick artist's controller ES, process corresponding adjustment strategy, if inconsistent or run into new failure symptom and cannot find matched rule in knowledge base, may there is exceptional fault in declarative procedure, starts the exceptional fault diagnostic method in inference machine, finally, diagnostic result exports user interface to by interpre(ta)tive system.If diagnostic result and actual testing result through exceptional fault diagnostic method are not inconsistent, the existing knowledge in declarative knowledge storehouse can not meet the requirement of diagnosis, need to supplement new knowledge, proceed to fresh information and obtain program, after supplementing new information, restart diagnostic routine, to be issued to the object of correct diagnosis in newly-added information support, thus the accuracy of assurance system to exceptional fault diagnosis.
The method according to this invention, can carry out reasoning from logic to situation in the middle of industrial processes, judgement industrial processes unusual condition and Change and Development rule, can more effectively to the fault-signal of commercial production dynamic process, effectively process, mutual relationship between fault analysis and judgement signal and reason, avoid the misjudgement of breaking down and the phenomenon of failing to judge, improve the intelligent and diagnosis efficiency of system, ensure stably normally production of industrial processes.
By in conjunction with the following drawings, to read after the detailed description of embodiment of the present invention, other features of the present invention, feature and advantage will become apparent.
Accompanying drawing explanation
Fig. 1 is according to the integrated method for diagnosing faults schematic flow diagram of industrial processes intelligence of the present invention;
Fig. 2 obtains Bayesian network model Establishing process figure in the integrated method for diagnosing faults of Fig. 1 industrial processes intelligence;
Fig. 3 obtains FTA and the integrated diagnosing and analyzing of FMEA model and the process flow diagram of processing in the integrated method for diagnosing faults of Fig. 1 industrial processes intelligence;
Fig. 4 obtains neural network model Analysis on Fault Diagnosis and processing flow chart in the integrated method for diagnosing faults of Fig. 1 industrial processes intelligence;
Fig. 5 obtains expert system Analysis on Fault Diagnosis and processing flow chart in the integrated method for diagnosing faults of Fig. 1 industrial processes intelligence;
Fig. 6 schematically shows a kind of industrial processes intelligence integrated trouble-shooter---field bus control system FCS, and Fig. 1 can realize to the whole bag of tricks shown in Fig. 5 in this system.
Embodiment
Below in conjunction with accompanying drawing, the specific embodiment of the present invention is elaborated.
Fig. 1 is according to the integrated method for diagnosing faults schematic flow diagram of industrial processes intelligence of the present invention.
First, in step 101, determine tested industrial processes object, at step 102 pair tested industrial processes object, carry out signals collecting, obtain relevant information; In step 103 pair industrial processes characteristics of objects information, process, state identifies; The process status parameter providing in step 105 integrating step 104 knowledge bases and database information disposal system, according to intelligent integrated-type method for diagnosing faults to data analysis and processing; In step 106 pair fault, identify; In step 107 pair fault, accurately judge and locate; In step 108, according to interpretation process system, diagnose decision-making, system is regulated effectively.
Fig. 2 is the Establishing process figure that obtains Bayesian network model in the integrated method for diagnosing faults of Fig. 1 industrial processes intelligence.
In step 201, model is set up and is started.
In step 202, be according to the cause-effect relationship of procedure parameter variable, determine topology of networks, variable can be chosen discrete type or continuous distribution type.
In step 203, it is Bayesian network assignment.For root node, determine its prior probability; For other node, to determine conditional probability.Determining of prior probability and conditional probability, can obtain according to expert info and Test Information.
In step 204, determine the causal relation between node, design conditions probability distribution table, it is the probability distribution based on even higher level of node different conditions.Determine after the prior probability of root node and the conditional probability of all the other nodes, just can determine according to bayesian theory and carry out on this basis statistical inference by the unconditional prior probability of all nodes.
In step 205, determine the probability distribution table of harsh degree.For the probability distribution of harsh degree, be revised according to test situation and to deepening continuously of understanding systematicly, to reflect the actual conditions of system.
In step 206, be belief propagation and deduction.This process is mainly to merge new Test Information and expert info, improves network structure node and probability distribution, and carries out last statistical inference.
In step 207, finish.
By belief propagation, constantly adjust more prior probability, conditional probability and the severity probability of node, realize the learning functionality of Bayesian network, this is the basis of carrying out statistical inference.
Determined after new prior probability, conditional probability and severity probability distribution table, just can utilize Bayesian network to carry out statistical inference.
Fig. 3 obtains FTA and the integrated diagnosing and analyzing of FMEA model and the process flow diagram of processing in the integrated method for diagnosing faults of Fig. 1 industrial processes intelligence.
In step 301, first gather crucial exceptional fault information and carry out FTA analysis as top event, form fault tree.
In step 302, by top event, find intermediate event.
In step 303, determine bottom event.
In step 304, the form expression way according to the expressed logical relation of fault tree in Bayesian network, is converted into this fault tree the FTA topological model of Bayesian network, determines important bottom event.
In Bayesian network model after conversion, according to each bottom event prior probability distribution of having supposed, utilize the information that in fault tree, certain event has occurred, analyze the conditional probability that each event occurs, and the bottom event of the maximum probability of selecting to break down is as important bottom event.
In step 305, the FMEA model of operation Bayesian network.
In step 306, around important bottom event, launch FMEA and analyze, provide severity tier definition table, and analyzing failure cause, fault mode, severity and fault effects, FMEA analysis result is converted into CFE Bayesian network model.Wherein, the failure cause of FMEA, fault mode and fault effects correspond respectively to reason layer, fault layer and the impact layer in CFE model.
In step 307, whether failure judgement clue exists, if existed, flow process forwards step 308 to, otherwise flow process forwards step 310 to
In step 308, the probability occurring according to each fault mode of fault tree analysis gained, utilize the CFE Bayesian network model of step 306 and causality and the level between the system failure, calculate the probability of happening of each fault mode and the probability that the combination of most probable fault mode occurs, determine the module most possibly break down, by localization of fault to module and propose innovative approach.
In step 309, use neural network expert system diagnostic techniques to carry out fault locating analysis.
In step 310, supplement the potential impact of bottom event, set up new fault tree pattern analysis, repeat above analytic process, flow process forwards step 301 to.
In step 311, flow process finishes, and realizes localization of fault, completes failure diagnostic process.
For industrial processes, whole system can be divided into a plurality of parts of different levels conventionally, comprising: reason layer analysis, fault layer analysis, affect layer analysis.In general, need to be the analysis result of the low level FMEA of system (Failure Modes and Effects Analysis) comprehensively to a high hierarchical structure, just can obtain FMEA analysis result generally, between analysis result at all levels, exist certain cause-effect relationship, that is: the impact of the fault mode of a low hierarchical system on a high level is exactly the fault mode of a high hierarchical system; And a low hierarchical system causes the fault mode of this fault effects, be the failure cause that high this fault of a hierarchical system is hit formula, be pushed into whole system on thus.By this iterative relation, the analysis result of a low hierarchical system can be brought into the analysis of a high hierarchical system.
By the trouble analysis system of Bayesian network, determined the mutual relationship between variable, making becomes possibility to analysis of complex system, carries out the FMEA research of complication system with Bayesian network, can merge various source-informations, under condition of uncertainty, carry out reasoning, improved confidence level.
Around important bottom event, launch FMEA and analyze, provide severity tier definition table, and analyzing failure cause, fault mode, severity and fault effects, FMEA analysis result is converted into CFE Bayesian network model.Wherein, the failure cause of FMEA, fault mode and fault effects correspond respectively to reason layer, fault layer and the impact layer in CFE model.
The diagnosis decision-making of CFE type Bayesian network be in conjunction with the prior probability of each node of reason layer and all fault node layers at its parent node to the conditional probability under stable condition, according to total probability formula reasoning must be out of order layer each node Prior Probability, when given sign information, by Bayesian formula, further revise this Prior Probability, obtain posterior probability values, thereby can determine the possibility that a certain node failure occurs.
Fig. 4 obtains neural network model Analysis on Fault Diagnosis and processing flow chart in the integrated method for diagnosing faults of Fig. 1 industrial processes intelligence.
In step 401, carry out system initialization.
In step 402, give fixed system input value and output expectation desired value.
In step 403, according to input and desired value, judge whether system has been learnt, if do not had, flow process goes to step 404, if completed, flow process enters duty, goes to step 408.
In step 404, calculate hidden layer and each cell value of output layer.
In step 405, calculate the deviation E of desired value and real output value.
In step 406, whether judgment bias E meets technological requirement, if do not met, flow process goes to step 410, if met, flow process goes to step 407.
In step 407, judge whether the deviation e of all unit meets technological requirement, if do not met, flow process goes to step 410, if met, flow process goes to step 408.
In step 408, ask output vector.
In step 409, start expert system and provide diagnostic result.
In step 410, calculate the error of hidden layer unit.
In step 411, regularized learning algorithm rate.
In step 412, adjust middle layer to the connection weights of output layer and the threshold value of output layer unit.
In step 413, adjust input layer to the connection weights in middle layer and the threshold value of middle layer elements.
In step 414, study completes.
This neural net method is first to learn, and after having learnt, starts expert system, then provides diagnostic result, and its study is to have directed learning, and people is for providing input and output sample.
Fig. 5 obtains expert system Analysis on Fault Diagnosis and processing flow chart in the integrated method for diagnosing faults of Fig. 1 industrial processes intelligence.
In step 501, start expert system.
In step 502, operation expert's priori storehouse.
In step 503, search case library carries out reasoning, determines malfunctioning module, obtains the diagnostic result based on reasoning by cases, as diagnostic result is unsatisfied with or is come to nothing, then search rule storehouse.
In step 504, search rule storehouse, obtains the diagnostic result of rule-based reasoning; As still diagnostic result being unsatisfied with or being come to nothing, then carry out model reasoning, obtain the diagnostic result based on model reasoning.
In step 505, based on model, carry out reasoning.
In step 506, the diagnostic result that different inference methods are obtained is comprehensively analyzed.
In step 507, operation interpretation process system, the confirmation result that reasoning is obtained makes an explanation.
In step 508, diagnostic result output.
In step 509, finish.
Fig. 6 schematically shows a kind of industrial processes intelligence integrated trouble-shooter---field bus control system FCS, and Fig. 1 can realize to the whole bag of tricks shown in Fig. 5 in this system.
601OS represents operator's computing machine, 602ES represents slip-stick artist's controller, 603SCS represents fault diagnosis supervisory control comuter, 604CG represents computing machine gateway, 605SNET represents monitor network, and 606FBI represents field-bus interface, 607FNET fieldbus networks, 608 represent FD fieldbus instrument and equipment, and 609 represent industrial processes.
OS operator's computing machine 601, monitors, operates and manage production run for site technique operator, carries out man-machine conversation.
ES slip-stick artist's controller 602, carries out system generation for field engineer to FCS and fault diagnosis is safeguarded, the establishment of control engineering Shi Jinhang control loop configuration programming and special applications software is provided.
SCS fault diagnosis supervisory control comuter 603, for foundation and the updating maintenance of foundation, database and the knowledge base of fault diagnosis model, production run is carried out to fault diagnosis, forecast and analysis, guarantees safety in production.
CG computing machine gateway 604, for connecting monitor network 605 and production management network, realizes the intercommunication mutually between them.
FBI field-bus interface 606, under connect FNET fieldbus networks 607, on connect SNET monitor network 605, thereby realize interconnecting of FNET fieldbus networks 607 and SNET monitor network 605.
FD fieldbus instrument and equipment 608, include: sensor, transmitter, actuator, instrument power, power supply adaptor and safety barrier etc., be used for input, output, the computing of the data-signal of fault diagnosis, the adjusting of fault, control and communicate by letter, and functional block is provided, to form control loop in bus at the scene.
Technical scheme: when trouble-shooter field bus control system FCS moves, first by FD fieldbus instrument and equipment 608, gather industrial process fault diagnosis object 609 necessary diagnostic message and failure symptoms, by FNET fieldbus 607, FBI field-bus interface 606 and SNET monitor network 605, information is uploaded to operator's computing machine 601, slip-stick artist's controller 602 and fault diagnosis supervisory control comuter 603, and by CG computing machine gateway 604, carry out the intercommunication of monitor network and production management network.By FCS system, according to diagnostic knowledge, utilize intelligent integrated model for fault diagnosis to carry out reasoning, according to the bottom fault knowledge being stored on network weight, complete to detecting the fault analysis of data, with the consequent output of failure symptom, inference machine receives after the bottom fault diagnosis result of network diagnostic systems, in conjunction with the target level knowledge in knowledge base, strategy carries out reasoning diagnosis by inference, obtain the prediction conclusion of high-rise fault diagnosis, if prediction fault is identified, by elementary knowledge, instruct inference machine from prediction conclusion, utilize target level knowledge by inference strategy carry out reasoning diagnosis, if the failure symptom that reasoning draws is consistent with the failure symptom of network diagnosis model, high-rise fault is identified and is sent to interpretation process system, interpretation process system points out according to elementary knowledge the position that reason that fault may occur and fault occur, and provide the correlation method of handling failure, by slip-stick artist's controller ES602, process corresponding adjustment strategy, if inconsistent or run into new failure symptom and cannot find matched rule in knowledge base, may there is exceptional fault in declarative procedure, starts the exceptional fault diagnostic method in inference machine, finally, diagnostic result exports user interface to by interpre(ta)tive system.If diagnostic result and actual testing result through exceptional fault diagnostic method are not inconsistent, the existing knowledge in declarative knowledge storehouse can not meet the requirement of diagnosis, need to supplement new knowledge, proceed to fresh information and obtain program, after supplementing new information, restart diagnostic routine, to be issued to the object of correct diagnosis in newly-added information support, thus the accuracy of assurance system to exceptional fault diagnosis.
If different inference methods obtain similar diagnostic result, diagnostic result is more credible; If the diagnostic result that different inference methods obtain differs larger, need diagnostic result to make a concrete analysis of, at this moment can select the diagnostic result based on model, or the larger diagnostic result of confidence level is as diagnosis.
At Fig. 1, to the basis of the process flow diagram shown in Fig. 5, in conjunction with native system technician, without creationary work, can develop more application software, carry out production run fault diagnosis.
Beneficial effect and the advantage of intelligence integrated diagnosis method: level of integrated system is high, can give full play to the advantage of various diagnostic methods, overcome the deficiency that method exists separately, and can find to greatest extent way to solve the problem, fault is accurately located, avoid the misjudgement of fault and the phenomenon of failing to judge, there is higher diagnosis efficiency and diagnosis rate.
Claims (8)
1. the integrated method for diagnosing faults of industrial processes intelligence, is characterized in that, comprises the following steps:
(1) detection of industrial process fault diagnosis data, signals collecting;
(2) according to the signal collecting, carry out analysis and the processing of characteristics of objects;
(3), according to characteristics of objects, according to the integrated method for diagnosing faults of intelligence, to production run Analysis on Fault Diagnosis, fault is identified;
Step (3) realizes by following steps:
A) foundation of Bayesian network model;
B) comprehensive analysis and the processing of FTA and FMEA model;
C) neural network expert system Analysis on Fault Diagnosis and processing;
(4) trouble-shooting reason, carries out fault and accurately locates and diagnose decision-making, effectively regulates production run;
The diagnostic procedure of the integrated method for diagnosing faults of intelligence: whole network diagnostic systems consists of Bayesian network model, FTA and FMEA failure effect analysis (FEA) model and neural network model, in order to complete foundation, fault analysis and the reasoning from logic process of model; Here introducing Bayesian network model is in order to solve the uncertain factor of industrial processes, improves the accuracy of industrial process fault diagnosis; Adopt FTA and FMEA technology, can help system determine in early days in fault diagnosis the module that causes software failure, dwindle the scope of localization of fault; In addition, Bayesian network technology has very large similarity with FTA and FMEA in inference mechanism and system state, simultaneously can also improve their descriptive power, the probabilistic relation between, random occurrence uncertain by expressing obtains how useful conclusion, for fault diagnosis and location provides foundation;
Here FMEA Failure Mode Effective Analysis method be each module in analytic system, a kind of inductive method likely affecting setting up all issuable fault modes and system is caused; FTA fault tree analysis is the logical relation for showing which module of system has fault, external event or their combination to cause system to break down; Use separately FTA and FMEA all respectively to have drawback, only utilize FMEA can strengthen workload, and easily omit fault mode and fault effects, only utilize FTA easily to omit again the bottom event that causes top event to occur; Therefore, be the drawback of avoiding using separately FTA and FMEA to bring, FTA and FMEA are combined with Bayesian network technology respectively, turn to bayesian network structure model and carry out fault diagnosis, can improve fault diagnosis efficiency;
First diagnostic model is analyzed phenomenon of the failure by the FTA based on Bayesian network and FMEA Comprehensive Analysis Technique, fault Primary Location is arrived to a certain module, then use neural network expert system diagnostic techniques to realize fault and further accurately locate, complete failure diagnostic process;
For industrial processes, whole system can be divided into a plurality of parts of different levels conventionally, comprising: reason layer analysis, fault layer analysis, affect layer analysis; In general, need to be the analysis result of the low level FMEA of system (Failure Modes and Effects Analysis) comprehensively to a high hierarchical structure, just can obtain FMEA analysis result generally, between analysis result at all levels, exist certain cause-effect relationship, that is: the impact of the fault mode of a low hierarchical system on a high level is exactly the fault mode of a high hierarchical system; And a low hierarchical system causes the fault mode of this fault effects, be the failure cause that high this fault of a hierarchical system is hit formula, be pushed into whole system on thus; By this iterative relation, the analysis result of a low hierarchical system can be brought into the analysis of a high hierarchical system;
By the trouble analysis system of Bayesian network, determined the mutual relationship between variable, making becomes possibility to analysis of complex system, carries out the FMEA research of complication system with Bayesian network, can merge various source-informations, under condition of uncertainty, carry out reasoning, improved confidence level;
Around important bottom event, launch FMEA and analyze, provide severity tier definition table, and analyzing failure cause, fault mode, severity and fault effects, FMEA analysis result is converted into CFE Bayesian network model; Wherein, the failure cause of FMEA, fault mode and fault effects correspond respectively to reason layer, fault layer and the impact layer in CFE model;
The diagnosis decision-making of CFE type Bayesian network be in conjunction with the prior probability of each node of reason layer and all fault node layers at its parent node to the conditional probability under stable condition, according to total probability formula reasoning must be out of order layer each node Prior Probability, when given sign information, by Bayesian formula, further revise this Prior Probability, obtain posterior probability values, thereby can determine the possibility that a certain node failure occurs;
Adopt neural network by elementary knowledge being learnt the bottom fault of processing procedure, target level knowledge and the high-rise diagnosing malfunction of intergrade knowledge to process that expert system stores according to knowledge base;
This neural net method is first to learn, and after having learnt, starts expert system, then provides diagnostic result, and its study is to have directed learning, and people is for providing input and output sample;
According to intelligent integrated approach, according to characteristics of objects, carry out production run Analysis on Fault Diagnosis, fault is identified, and trouble-shooting reason, carries out fault and accurately locates, and diagnose decision-making, effectively carry out system adjusting, thereby industrial processes can be carried out smoothly;
Whole integrated system makes full use of model internal fault diagnostic method advantage separately and carries out complementation, thereby effectively overcome the deficiency that they exist separately, has good diagnosis efficiency and reliability.
2. the integrated method for diagnosing faults of a kind of industrial processes intelligence according to claim 1, the foundation that it is characterized in that Bayesian network model is to realize by carrying out following steps:
(1) according to the cause-effect relationship of procedure parameter variable, determine topology of networks;
(2) Bayesian network assignment, for root node, determine its prior probability, for other node, will determine conditional probability, and determining of prior probability and conditional probability, can obtain according to expert info and Test Information;
(3) determine the causal relation between node, design conditions probability distribution table, it is the probability distribution based on even higher level of node different conditions; Determine after the prior probability of root node and the conditional probability of all the other nodes, just can determine according to bayesian theory and carry out on this basis statistical inference by the unconditional prior probability of all nodes;
(4) determine the probability distribution table of harsh degree, for the probability distribution of harsh degree, be revised according to test situation and to deepening continuously of understanding systematicly, to reflect the actual conditions of system;
(5) belief propagation and deduction, this process is mainly to merge new Test Information and expert info, improves network structure node and probability distribution, and carries out last statistical inference; Pass through belief propagation, constantly adjust more prior probability, conditional probability and the severity probability of node, realize the learning functionality of Bayesian network, this is the basis of carrying out statistical inference, determined after new prior probability, conditional probability and severity probability distribution table, just can utilize Bayesian network to carry out statistical inference.
3. the integrated method for diagnosing faults of a kind of industrial processes intelligence according to claim 1, is characterized in that the integrated diagnosing and analyzing of FTA and FMEA model and processing are to realize by carrying out following steps:
(1) first gather crucial exceptional fault information and carry out FTA analysis as top event, form fault tree;
(2) by top event, find intermediate event;
(3) determine bottom event, form expression way according to the expressed logical relation of fault tree in Bayesian network, this fault tree is converted into the FTA topological model of Bayesian network, determine important bottom event, in Bayesian network model after conversion, according to each bottom event prior probability distribution of having supposed, utilize the information that in fault tree, certain event has occurred, analyze the conditional probability that each event occurs, and the bottom event of the maximum probability of selecting to break down is as important bottom event;
(4) the FMEA model of operation Bayesian network;
(5) fault clue existence judgement, if existed, arrives module by localization of fault; If there is no, supplement the potential impact of bottom event, set up new fault tree pattern analysis;
(6) probability occurring according to each fault mode of fault tree analysis gained, utilize causality and level between CFE Bayesian network model and the system failure, calculate the probability of happening of each fault mode and the probability that the combination of most probable fault mode occurs, determine the module most possibly breaking down;
(7) use neural network expert system diagnostic techniques to carry out fault locating analysis, by localization of fault to module and propose innovative approach;
(8) supplement the potential impact of bottom event, set up new fault tree pattern analysis, continue to repeat above-mentioned analytic process.
4. the integrated method for diagnosing faults of a kind of industrial processes intelligence according to claim 1, is characterized in that the Analysis on Fault Diagnosis of neural network model and processing are to realize by carrying out following steps:
(1) first carry out system initialization;
(2) give fixed system input value and output expectation desired value;
(3) according to input and desired value, judge whether systematic learning completes, if do not completed, go to and calculate hidden layer and each cell value of output layer, if completed, enter duty, ask output vector;
(4) calculate hidden layer and each cell value of output layer;
(5) calculate the deviation E of desired value and real output value;
(6) deviation E judges whether to meet technological requirement, if do not met, calculates the error of hidden layer unit; If met, judge whether the deviation e of all unit meets technological requirement, if do not met, calculate the error of hidden layer unit; If meet, ask output vector;
(7) ask output vector;
(8) start expert system and provide diagnostic result;
(9) calculate the error of hidden layer unit;
(10) regularized learning algorithm rate;
(11) adjust middle layer to the connection weights of output layer and the threshold value of output layer unit;
(12) adjust input layer to the connection weights in middle layer and the threshold value of middle layer elements;
(13) complete learning process.
5. the integrated method for diagnosing faults of a kind of industrial processes intelligence according to claim 1, is characterized in that expert system Analysis on Fault Diagnosis and processing procedure are to realize by carrying out following steps:
(1) start expert system;
(2) operation expert's priori storehouse;
(3) search case library carries out reasoning, determines malfunctioning module, obtains the diagnostic result based on reasoning by cases, as diagnostic result is unsatisfied with or is come to nothing, then search rule storehouse;
(4) search rule storehouse, obtains the diagnostic result of rule-based reasoning; As still diagnostic result being unsatisfied with or being come to nothing, then carry out model reasoning, obtain the diagnostic result based on model reasoning;
(5) based on model, carry out reasoning;
(6) diagnostic result different inference methods being obtained is comprehensively analyzed;
(7) operation interpretation process system, the confirmation result that reasoning is obtained makes an explanation;
(8) diagnostic result output, if different inference method obtains similar diagnostic result, diagnostic result is more credible; If the diagnostic result that different inference methods obtain differs larger, need diagnostic result to make a concrete analysis of, at this moment can select the diagnostic result based on model, or the larger diagnostic result of confidence level is as diagnosis.
6. according to the integrated method for diagnosing faults of a kind of industrial processes intelligence described in any one in claim 1 to 5, it is characterized in that forming integrated troubleshooting step, is an organic whole.
7. the integrated method for diagnosing faults of intelligence according to claim 6, is characterized in that diagnosis object is industrial processes.
8. the integrated trouble-shooter of industrial processes intelligence, for realizing the integrated method for diagnosing faults of intelligence as described in claim 1-7, is characterized in that comprising:
FD fieldbus instrument and equipment, include: sensor, transmitter, actuator, instrument power, power supply adaptor and safety barrier etc., for the production of input, output, computing, the control of process failure diagnosis data-signal with communicate by letter, and functional block is provided, to form control loop in bus at the scene;
FNET fieldbus networks, under connect FD fieldbus instrument and equipment, on connect FBI field-bus interface, carry out the mutual exchange of on-site signal and control signal;
FBI field-bus interface, under connect FNET fieldbus, on connect SNET monitor network, thereby realize interconnecting of FNET fieldbus and SNET monitor network;
SNET monitor network, for connecting FBI field-bus interface, OS operator's computing machine, ES slip-stick artist's controller, SCS fault diagnosis supervisory control comuter and CG computing machine gateway, carries out signal transmission;
OS operator's computing machine, monitors, operates and manage production run for site technique operator, carries out man-machine conversation;
ES slip-stick artist's controller, carries out system generation for field engineer to FCS and fault diagnosis is safeguarded, the establishment of control engineering Shi Jinhang control loop configuration programming and special applications software is provided;
SCS fault diagnosis supervisory control comuter, for foundation and the updating maintenance of foundation, database and the knowledge base of fault diagnosis model, production run is carried out to fault diagnosis, forecast and analysis, guarantees safety in production;
CG computing machine gateway, for connecting monitor network and production management network, realizes the intercommunication mutually between them.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201110035484.XA CN102637019B (en) | 2011-02-10 | 2011-02-10 | Intelligent integrated fault diagnosis method and device in industrial production process |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201110035484.XA CN102637019B (en) | 2011-02-10 | 2011-02-10 | Intelligent integrated fault diagnosis method and device in industrial production process |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN102637019A CN102637019A (en) | 2012-08-15 |
| CN102637019B true CN102637019B (en) | 2014-04-02 |
Family
ID=46621436
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201110035484.XA Expired - Fee Related CN102637019B (en) | 2011-02-10 | 2011-02-10 | Intelligent integrated fault diagnosis method and device in industrial production process |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102637019B (en) |
Families Citing this family (61)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| DE102012014584A1 (en) * | 2012-07-23 | 2014-01-23 | Hottinger Baldwin Messtechnik Gmbh | Measured variable sensor with internal data memory |
| CN103001812B (en) * | 2013-01-06 | 2015-12-02 | 国网山西省电力公司信息通信分公司 | Intelligent electric power communication failure diagnostic system |
| CN103336198B (en) * | 2013-06-21 | 2015-06-17 | 中国人民解放军国防科学技术大学 | Electrical system fault diagnosing device |
| CN103810383A (en) * | 2014-01-27 | 2014-05-21 | 中国航天标准化研究所 | Method for identifying and analyzing single-point failure mode of launch vehicle |
| CN104238505A (en) * | 2014-08-28 | 2014-12-24 | 华南理工大学 | Quality control expert system and method for FPC manufacturing process |
| CN104331058A (en) * | 2014-11-24 | 2015-02-04 | 重庆科技学院 | Fault diagnosis ontology construction method |
| CN105717887B (en) * | 2014-12-05 | 2018-03-02 | 克拉玛依红有软件有限责任公司 | Oil vapor treatment station intelligent early-warning system based on Internet of Things sensing data |
| CN104363129B (en) * | 2014-12-10 | 2017-08-04 | 国家电网公司 | A kind of network event association analysis and dynamic early-warning method |
| CN104865956B (en) * | 2015-03-27 | 2017-07-07 | 重庆大学 | Sensor fault diagnosis method in a kind of complication system based on Bayesian network |
| CN104991549B (en) * | 2015-05-27 | 2017-06-30 | 中国铁路总公司 | Track circuit red band method for diagnosing faults based on FTA and multi-layer fuzzy nerve sub-network |
| CN105138802B (en) * | 2015-09-25 | 2018-01-12 | 北京理工大学 | A kind of gun barrel intelligent design system and design method |
| CN108139728B (en) * | 2015-10-09 | 2021-09-28 | 费希尔-罗斯蒙特系统公司 | System and method for representing a cause and effect matrix as a set of numerical representations |
| CN105467953B (en) * | 2015-11-11 | 2018-01-09 | 中国科学院软件研究所 | A kind of representation of knowledge and its automation application method towards industrial big data |
| CN106249599B (en) * | 2016-09-28 | 2020-01-10 | 河南理工大学 | Neural network prediction-based networked control system fault detection method |
| CN106529582B (en) * | 2016-10-24 | 2019-08-20 | 杭州电子科技大学 | For the prior probability appraisal procedure for introducing expert opinion in Bayesian network |
| CN106529581B (en) * | 2016-10-24 | 2019-08-20 | 杭州电子科技大学 | A kind of bridge crane method for diagnosing faults based on Bayesian network |
| CN106547967B (en) * | 2016-11-01 | 2020-07-28 | 哈尔滨工程大学 | Diesel engine fuel system maintenance decision method based on cost analysis |
| CN106778210B (en) * | 2016-12-16 | 2020-04-07 | 成都巧班科技有限公司 | Industrial control system function safety verification method based on immune learning |
| CN108241114B (en) * | 2016-12-26 | 2020-05-26 | 北京邮电大学 | Online fault detection method and device for hybrid circuit chip and fault detection circuit |
| CN106959662B (en) * | 2017-05-10 | 2019-03-08 | 东北大学 | A kind of identification of electric melting magnesium furnace unusual service condition and control method |
| CN107193275A (en) * | 2017-07-01 | 2017-09-22 | 浙江三青环保科技有限公司 | A kind of equipment operation intelligent diagnosis system and diagnostic method |
| CN107357228A (en) * | 2017-07-11 | 2017-11-17 | 中兴耀维科技江苏有限公司 | A kind of method for monitoring fault location failure analysis and control on-line |
| CN107562034A (en) * | 2017-07-14 | 2018-01-09 | 宝沃汽车（中国）有限公司 | Fault handling method and processing system on line |
| CN107566170B (en) * | 2017-08-24 | 2020-07-14 | 广东电网有限责任公司珠海供电局 | Rule base-based automatic detection method for distribution network automatic telecontrol channel |
| CN108051637A (en) * | 2017-12-08 | 2018-05-18 | 国网河北省电力有限公司电力科学研究院 | A kind of Intelligent electric energy meter clock battery failures diagnostic method |
| CN108170566A (en) * | 2017-12-18 | 2018-06-15 | 新疆金风科技股份有限公司 | Product failure information processing method, system, equipment and collaboration platform |
| WO2019222152A1 (en) * | 2018-05-18 | 2019-11-21 | Siemens Aktiengesellschaft | Online fault localization in industrial processes without utilizing a dynamic system model |
| CN108803515A (en) * | 2018-06-20 | 2018-11-13 | 芜湖乐锐思信息咨询有限公司 | A kind of computer system for industry monitoring production procedure |
| JP7091174B2 (en) * | 2018-07-09 | 2022-06-27 | キヤノン株式会社 | System, system control method and program |
| CN109143851B (en) * | 2018-07-11 | 2021-06-01 | 佛山科学技术学院 | Method for recognizing multi-mark fault deep learning and intelligently expressing result thereof |
| CN109063839B (en) * | 2018-09-26 | 2020-08-14 | 北京航天自动控制研究所 | Complex symptom construction method of simulation temporal logic of expert system |
| CN109445406B (en) * | 2018-10-18 | 2021-05-18 | 西南交通大学 | Industrial control system safety detection method based on scene test and transaction search |
| CN109634233A (en) * | 2018-12-06 | 2019-04-16 | 南京邮电大学 | Industrial big data intellectual analysis decision-making technique, readable storage medium storing program for executing and terminal |
| CN109902373B (en) * | 2019-02-21 | 2023-06-23 | 国网山东省电力公司临沂供电公司 | Fault diagnosis and positioning method and system for district transformer substation |
| CN109894476B (en) * | 2019-02-28 | 2020-06-26 | 北京首钢股份有限公司 | Fault diagnosis method and device for hydraulic equipment of cold-rolled silicon steel production line |
| WO2020183340A1 (en) * | 2019-03-14 | 2020-09-17 | Abb Schweiz Ag | A method of detecting faults in intelligent electronic devices |
| EP3712721A1 (en) * | 2019-03-19 | 2020-09-23 | Siemens Aktiengesellschaft | Safety-related diagnosis reports |
| CN110083636A (en) * | 2019-04-04 | 2019-08-02 | 深圳市华星光电技术有限公司 | The detection system and detection method of panel defect |
| CN110146316A (en) * | 2019-04-10 | 2019-08-20 | 广州瑞松北斗汽车装备有限公司 | The fault diagnosis method and system of auto production line function counter |
| CN109902308B (en) * | 2019-04-11 | 2023-04-18 | 中国民航大学 | Diagnosis method, system and device for aviation safety event analysis |
| CN110084500B (en) * | 2019-04-19 | 2020-03-31 | 深圳市德塔防爆电动汽车有限公司 | Electric vehicle safety control method based on safety tree probability and safety importance degree and electric vehicle |
| US11209817B2 (en) | 2019-05-06 | 2021-12-28 | Argo AI, LLC | Method and system for real-time diagnostics and fault monitoring in a robotic system |
| CN111897304B (en) * | 2019-05-06 | 2023-12-05 | 埃尔构人工智能有限责任公司 | Method, apparatus and system for real-time diagnostics and fault monitoring in a machine system |
| CN110500709B (en) * | 2019-07-29 | 2021-04-20 | 七彩安科智慧科技有限公司 | Air conditioner running state online judging method |
| CN110942150A (en) * | 2019-11-09 | 2020-03-31 | 上海航天设备制造总厂有限公司 | Intelligent diagnosis method for forming failure of parts of selective laser melting equipment |
| CN110968072B (en) * | 2019-11-19 | 2021-08-17 | 朱彤 | Electrical automation equipment monitoring system and method based on artificial intelligence |
| CN111207484B (en) * | 2019-12-13 | 2021-01-19 | 浙江大学 | Central air-conditioning system fault diagnosis method based on object-oriented Bayesian network |
| CN111459130B (en) * | 2020-03-04 | 2022-12-20 | 辽宁工程技术大学 | Method for determining potential possibility of electric system fault |
| CN111651931B (en) * | 2020-05-19 | 2022-05-17 | 浙江大学 | Blast furnace fault diagnosis rule derivation method based on deep neural network |
| US11416326B2 (en) | 2020-08-28 | 2022-08-16 | Sap Se | Systems and methods for failure diagnosis using fault tree |
| CN112415894B (en) * | 2020-11-20 | 2021-09-10 | 中国矿业大学 | Active learning and BN (boron nitride) -based heavy medium coal preparation process safe operation control method |
| CN112819028B (en) * | 2020-12-29 | 2024-02-02 | 重庆大学 | Fault diagnosis method of medical fresh air system based on fault diagnosis model |
| CN112731889A (en) * | 2021-01-18 | 2021-04-30 | 太原理工大学 | Intelligent air regulation management system and regulation and control method thereof |
| CN113234874B (en) * | 2021-05-14 | 2022-01-11 | 浙江大学 | Blast furnace fault monitoring and self-healing control system based on dynamic Bayesian network |
| CN113721579A (en) * | 2021-07-08 | 2021-11-30 | 河北工业大学 | Loom fault diagnosis method based on fusion of expert system and neural network algorithm |
| CN113589797A (en) * | 2021-08-06 | 2021-11-02 | 上海应用技术大学 | Intelligent diagnosis method and system for coke oven vehicle operation fault |
| CN114064911B (en) * | 2021-09-30 | 2024-07-16 | 中国核电工程有限公司 | Expert knowledge base modeling method and system for intelligent diagnosis system of nuclear power plant |
| CN115906642B (en) * | 2022-11-28 | 2023-07-28 | 东莞科达五金制品有限公司 | Bearing production detection control method and device |
| CN117114102A (en) * | 2023-10-13 | 2023-11-24 | 江苏前景瑞信科技发展有限公司 | Transformer fault diagnosis method based on Bayesian network and fault tree |
| CN117784742B (en) * | 2024-02-27 | 2024-05-28 | 广州优百特科技有限公司 | Fault diagnosis method of fatty acid biological feed production monitoring system |
| CN118586860A (en) * | 2024-06-06 | 2024-09-03 | 浙江省自然资源战略研究中心(浙江省自然资源调查登记中心) | Homeland change investigation checking system and method |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1685664A (en) * | 2002-07-29 | 2005-10-19 | 鲍米勒系统工程有限公司 | Computer network with diagnosis computer nodes |
| CN101526586A (en) * | 2009-04-17 | 2009-09-09 | 上海电力学院 | Embedded remote state monitoring system of generating unit |
-
2011
- 2011-02-10 CN CN201110035484.XA patent/CN102637019B/en not_active Expired - Fee Related
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1685664A (en) * | 2002-07-29 | 2005-10-19 | 鲍米勒系统工程有限公司 | Computer network with diagnosis computer nodes |
| CN101526586A (en) * | 2009-04-17 | 2009-09-09 | 上海电力学院 | Embedded remote state monitoring system of generating unit |
Non-Patent Citations (4)
| Title |
|---|
| 基于多传感器信息融合的加工过程监控;赵中敏;《化工自动化及仪表》;20080610;第35卷(第3期);1-5，9 * |
| 朱大奇.航空电子设备故障诊断技术研究.《中国博士学位论文全文数据库 工程科技II辑》.2003,(第2期), |
| 航空电子设备故障诊断技术研究;朱大奇;《中国博士学位论文全文数据库 工程科技II辑》;20030228(第2期);C031-4 * |
| 赵中敏.基于多传感器信息融合的加工过程监控.《化工自动化及仪表》.2008,第35卷(第3期),1-5,9. |
Also Published As
| Publication number | Publication date |
|---|---|
| CN102637019A (en) | 2012-08-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102637019B (en) | Intelligent integrated fault diagnosis method and device in industrial production process | |
| WO2023227071A1 (en) | Multi-model fused avionic product health assessment method | |
| Cai et al. | Bayesian networks in fault diagnosis | |
| Zio et al. | A data-driven fuzzy approach for predicting the remaining useful life in dynamic failure scenarios of a nuclear system | |
| Evsukoff et al. | Recurrent neuro-fuzzy system for fault detection and isolation in nuclear reactors | |
| EP3497527B1 (en) | Generation of failure models for embedded analytics and diagnostics | |
| CN107346466A (en) | A kind of control method and device of electric power dispatching system | |
| CN109242147A (en) | Signal fused fan condition prediction technique based on Bp neural network | |
| CN101859128A (en) | Knowledge-based fault prediction expert system for complex milling machine tool | |
| US11604461B2 (en) | Method and apparatus for monitoring the condition of subsystems within a renewable generation plant or microgrid | |
| CN102662831B (en) | Method of diagnosis strategy optimization for fault tolerant system | |
| CN110296124A (en) | Remote failure diagnosis system and method based on expert system | |
| CN108957230B (en) | Power distribution network fault diagnosis method and system based on hierarchical timing sequence self-adaptive fuzzy H network | |
| CN116205265A (en) | Power grid fault diagnosis method and device based on deep neural network | |
| CN114299338A (en) | Fault prediction and health management system for system management and related method | |
| Di Maio et al. | Transient identification by clustering based on Integrated Deterministic and Probabilistic Safety Analysis outcomes | |
| Basher | Autonomous control of nuclear power plants | |
| Yao et al. | A fault diagnosis expert system based on aircraft parameters | |
| CN109447424B (en) | Risk propagation analysis algorithm for electric power information physical system considering risk state propagation | |
| US20220197260A1 (en) | Faulty Variable Identification Technique for Data-Driven Fault Detection Within A Process Plant | |
| KR20220089853A (en) | Method for Failure prediction and prognostics and health management of renewable energy generation facilities using machine learning technology | |
| Borth et al. | Probabilistic health and mission readiness assessment at system-level | |
| CN115237091A (en) | Electromechanical device fault tracing method and system | |
| Zhang et al. | Analysis and Design of Fault Prediction and Health Assessment System for the optoelectronic Equipment | |
| Xin et al. | Research on Intelligent Flexible Health Management Platform |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20140402 Termination date: 20150210 |
|
| EXPY | Termination of patent right or utility model |