CN102662142B - Prediction method for multi-parameter identification fault of power electronic circuit based on RVM-QNN - Google Patents
Prediction method for multi-parameter identification fault of power electronic circuit based on RVM-QNN Download PDFInfo
- Publication number
- CN102662142B CN102662142B CN201210067840.0A CN201210067840A CN102662142B CN 102662142 B CN102662142 B CN 102662142B CN 201210067840 A CN201210067840 A CN 201210067840A CN 102662142 B CN102662142 B CN 102662142B
- Authority
- CN
- China
- Prior art keywords
- circuit
- qnn
- parameter
- rvm
- power electronic
- 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
- 238000000034 method Methods 0.000 title claims abstract description 29
- 238000012549 training Methods 0.000 claims abstract description 23
- 238000013528 artificial neural network Methods 0.000 claims abstract description 21
- 238000012544 monitoring process Methods 0.000 claims abstract description 7
- 230000008569 process Effects 0.000 claims description 10
- 230000002950 deficient Effects 0.000 claims description 8
- 238000007519 figuring Methods 0.000 abstract 1
- 230000006870 function Effects 0.000 description 12
- 238000012706 support-vector machine Methods 0.000 description 2
- 230000008901 benefit Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000015556 catabolic process Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 230000007123 defense Effects 0.000 description 1
- 238000006731 degradation reaction Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 239000003814 drug Substances 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000005530 etching Methods 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 230000036541 health Effects 0.000 description 1
- 230000010365 information processing Effects 0.000 description 1
- 230000002045 lasting effect Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000013178 mathematical model Methods 0.000 description 1
- 210000005036 nerve Anatomy 0.000 description 1
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a prediction method for multi-parameter identification faults of a power electronic circuit based on RVM-QNN, comprising the following steps: (1) setting different parameter values for various components in the power electronic circuit and acquiring circuit performance parameters; (2) constructing a QNN neural network multi-parameter identifier by taking the circuit performance parameters and component parameters as a training sample; (3) obtaining time series of the circuit performance parameters at intervals; (4) predicting the circuit performance parameters by using a RVM algorithm and acquiring circuit performance parameters at a time in the future; (5) identifying and acquiring parameters of various components in the circuit by taking the predicted circuit performance parameters as input; and (6) evaluating fault conditions of a system according to the predicted parameters of various components in the circuit. The prediction method is capable of monitoring operation status of a circuit in real time, predicting parameters of various components in the power electronic circuit at a time in the future, evaluating fault conditions of the circuit in the future, and figuring out causes of faults.
Description
Technical field
The present invention relates to a kind of Fault prediction in power electronic circuit method, especially a kind of Power Electronic Circuit multiparameter identification of defective Forecasting Methodology based on RVM-QNN.
Background technology
Along with the continuous expansion of engineering real system scale, the complicacy of system, pathosis and non-linear also progressively raising, therefore higher to security of system and reliability requirement.Failure prediction refers to according to the past and present monitoring information of system, predicts the time that fault occurs or judges whether following some time etching system breaks down.Modern electronic equipment has been penetrated into national defense and military, industry, traffic, IT, agricultural, communication, business, medicine manufacture so that in each neighborhood systems such as household electrical appliance, the loss causing because of electronic failure also increases greatly, therefore, for the power-supply system effect of electronic equipment power supply is more important.Power Electronic Circuit is the core of power-supply system, study its failure prediction method, ensure that it is safe, lasting, reliably working is a part for electronic equipment health control, study hotspot and the difficult point in failure prediction field are become, at huge, the complex structure of investment, the field such as Aero-Space, nuclear energy that reliability requirement is high has Great significance.
The fault of Power Electronic Circuit is mainly caused by the degradation failure that respectively forms components and parts in circuit, therefore, if can dope each components and parts variation tendency in future that forms in circuit, determine the parameter value of following each components and parts of moment, can evaluate fault situation.
On the other hand, quantum nerve network is the product that quantum calculation combines with neural network, and it has better stability and validity, also has Fast Learning ability and very high information processing rate, can be used in the identification of Power Electronic Circuit real-time parameter.
Method Using Relevance Vector Machine (RVM) is the one sparse probability model similar to support vector machine (SVM) that Michael E.Tipping proposes, the training of RVM is carried out under Bayesian frame, can carry out Time Series Regression and estimate prediction.The advantage of RVM is can utilize Bayes's evidence process automatically to determine the super parameter of model in training process, and do not need to use crosscheck, use support vector is little, and reduces the computing time of export target amount predicted value, can be for the real-time estimate of Power Electronic Circuit.
Summary of the invention
Object of the present invention, be to provide a kind of Power Electronic Circuit multiparameter identification of defective Forecasting Methodology based on RVM-QNN, it can Real-Time Monitoring circuit operation conditions, each component parameter in predict future moment Power Electronic Circuit, and then assess following fault situation, and can judge failure cause.
In order to reach above-mentioned purpose, solution of the present invention is:
A Power Electronic Circuit multiparameter identification of defective Forecasting Methodology based on RVM-QNN, comprises the steps:
(1) each components and parts in Power Electronic Circuit are arranged to different parameter values, monitor the circuit measuring point signal in corresponding situation, and calculate acquisition cuicuit performance parameter;
(2) using the middle circuit performance parameters that obtained of step (1), component parameter as training sample, QNN neural network is trained, build QNN neural network multiparameter identifier;
(3) in circuit working process, at interval of a period of time, monitoring system measuring point signal, counting circuit performance parameter, the time series of acquisition cuicuit performance parameter;
(4) utilize RVM algorithm to predict circuit performance parameters, obtain the circuit performance parameters in following certain moment;
(5) taking the circuit performance parameters predicted as input, utilize the QNN neural network identifier after training in step (2), the each component parameter of circuit in following certain moment is obtained in identification;
(6) according to the each component parameter of circuit of predicting, system is carried out to fault state assessment.
In above-mentioned steps (2), while building QNN neural network multiparameter identifier, suppose that in predicted Power Electronic Circuit, components and parts number is N, component parameter number is also N, and circuit performance parameters is M, and the building process of identifier is:
(21) QNN neural network structure is three layers, and wherein input layer is M, and output layer node is N, and middle layer node is
round numbers;
(22) produce initialization weights by random function;
(23) use training sample to network training, by the actual output computational grid output error of desired output and network, by minimum weights and the quantum interval of adjusting each layer of network output error;
(24) will complete QNN network after training as multiparameter identifier.
In above-mentioned steps (4), utilize the method for RVM algorithm predicts circuit performance parameters as follows:
(41) performance parameter of circuit is X
i, i=1,2 ..., M, the performance parameter time series of circuit is X
i, 1, X
i, 2, X
i, 3..., obtain the performance parameter X of institute's prediction circuit
ithe performance parameter sequence in p moment is { X continuously
i, j, X
i, j+1..., X
i, j+p, j=1,2,3 ..., p=1,2,3,
(42) utilize RVM to { X
i, j, X
i, j+1..., X
i, j+pcarry out the prediction of forward direction q step, obtain the performance parameter X of (j+p+q) moment circuit
i, j+p+q, q=1,2,3 ...
In above-mentioned steps (5), the method that following certain each parameter of moment is obtained in identification is: with X
i, j+p+qfor input, i=1,2 ..., M, picks out each component parameter θ in (j+p+q) moment
r, j+p+q, r=1,2 ..., N.
Adopt after such scheme, the present invention is by obtaining the circuit performance parameters of Power Electronic Circuit under different component parameters, using circuit performance parameters, component parameter as training sample, off-line training QNN neural network, builds QNN neural network multiparameter identifier; By real time on-line monitoring Circuits System signal, the circuit performance parameters time series in acquisition cuicuit practical work process, utilizes RVM algorithm predicts to go out the circuit performance parameters in following certain moment; Finally, taking the circuit performance parameters in following certain moment as input, adopt QNN neural network identifier, obtain the each component parameter value of circuit in following certain moment.According to the each component parameter of moment in future picking out, system failure situation is assessed, and can be judged failure cause.The present invention, without the mathematical model of setting up Power Electronic Circuit, can realize the online failure prediction of Power Electronic Circuit.
Brief description of the drawings
Fig. 1 is process flow diagram of the present invention.
Embodiment
Below with reference to accompanying drawing, technical scheme of the present invention is elaborated.
As shown in Figure 1, the invention provides a kind of Power Electronic Circuit multiparameter identification of defective Forecasting Methodology based on RVM-QNN (Relevant Vector Machine-Quantum Neural Networks), comprise the steps:
(1) each components and parts in Power Electronic Circuit are arranged to different parameter values, monitor the circuit measuring point signal in corresponding situation, and calculate acquisition cuicuit performance parameter;
(2) using the middle circuit performance parameters that obtained of step (1), component parameter as training sample, QNN neural network is trained, build QNN neural network multiparameter identifier;
While building QNN neural network multiparameter identifier, suppose that in predicted Power Electronic Circuit, components and parts number is N, component parameter number is also N system, and circuit performance parameters is M (as output average voltage, electric current etc.), and the building process of identifier is:
(21) QNN neural network structure is three layers, and wherein input layer is M, and output layer node is N, and middle layer node is
(round numbers);
(22) produce initialization weights by random function.Random function is the function that produces number, for generation of random number, is prior art, can adopt the rand function in Matlab to produce by equally distributed random number between (0,1) in the present invention.
(23) use training sample to network training, by the actual output computational grid output error of desired output and network, by minimum weights and the quantum interval of adjusting each layer of network output error;
(24) will complete QNN network after training as multiparameter identifier.
(3) in circuit working process, at interval of a period of time, monitoring system measuring point signal, counting circuit performance parameter, the time series of acquisition cuicuit performance parameter;
(4) utilize RVM algorithm to predict circuit performance parameters, obtain the circuit performance parameters in following certain moment;
It is as follows that described RVM algorithm relevant arranges:
A) by Gaussian wavelet basis function structure RVM kernel function:
The Gaussian wavelet basis function φ (t) being made up of the second derivative of Gaussian function, is shown in following formula.
Wherein, t is time variable.
By above formula Gaussian wavelet basis function structure RVM kernel function, see following formula.
Wherein, x is all training samples, x
ibe i training sample, d is training sample number, and a, σ are kernel functional parameter, and a is kernel function width.
B) kernel functional parameter a=0.6, σ=1.
Utilize the method for RVM algorithm predicts circuit performance parameters as follows:
(41) performance parameter of circuit is X
i, i=1,2 ..., M, the performance parameter time series of circuit is X
i, 1, X
i, 2, X
i, 3..., obtain the performance parameter X of institute's prediction circuit
ithe performance parameter sequence in p moment is { X continuously
i, j, X
i, j+1..., X
i, j+p, j=1,2,3 ..., p=1,2,3,
(42) utilize RVM to { X
i, j, X
i, j+1..., X
i, j+pcarry out the prediction of forward direction q step, obtain the performance parameter X of (j+p+q) moment circuit
i, j+p+q, q=1,2,3 ...
(5) taking the circuit performance parameters predicted as input, utilize the QNN neural network identifier after training in step (2), the each component parameter of circuit in following certain moment is obtained in identification;
Specifically, with X
i, j+p+qfor input (i=1,2 ..., M), pick out each component parameter θ in (j+p+q) moment
r, j+p+q, r=1,2 ..., N.
(6) according to the each component parameter of circuit of predicting, system is carried out to valuation condition evaluation.
Above embodiment only, for explanation technological thought of the present invention, can not limit protection scope of the present invention with this, every technological thought proposing according to the present invention, and any change of doing on technical scheme basis, within all falling into protection domain of the present invention.
Claims (3)
1. the Power Electronic Circuit multiparameter identification of defective Forecasting Methodology based on RVM-QNN, is characterized in that comprising the steps:
(1) each components and parts in Power Electronic Circuit are arranged to different parameter values, monitor the circuit measuring point signal in corresponding situation, and calculate acquisition cuicuit performance parameter;
(2) using the middle circuit performance parameters that obtained of step (1), component parameter as training sample, QNN neural network is trained, build QNN neural network multiparameter identifier; While building QNN neural network multiparameter identifier, suppose that in predicted Power Electronic Circuit, components and parts number is N, component parameter number is also N, and circuit performance parameters is M, and the building process of identifier is:
(21) QNN neural network structure is three layers, and wherein input layer is M, and output layer node is N, and middle layer node is
round numbers;
(22) produce initialization weights by random function;
(23) use training sample to network training, by the actual output computational grid output error of desired output and network, by minimum weights and the quantum interval of adjusting each layer of network output error;
(24) will complete QNN network after training as multiparameter identifier;
(3) in circuit working process, at interval of a period of time, monitoring system measuring point signal, counting circuit performance parameter, the time series of acquisition cuicuit performance parameter;
(4) utilize RVM algorithm to predict circuit performance parameters, obtain the circuit performance parameters in following certain moment;
(5) taking the circuit performance parameters predicted as input, utilize the QNN neural network identifier after training in step (2), the each component parameter of circuit in following certain moment is obtained in identification;
(6) according to the each component parameter of circuit of predicting, system is carried out to fault state assessment.
2. a kind of Power Electronic Circuit multiparameter identification of defective Forecasting Methodology based on RVM-QNN as claimed in claim 1, is characterized in that: the method for utilizing RVM algorithm to predict circuit performance parameters in described step (4) is as follows:
(41) performance parameter of circuit is X
i, i=1,2 ..., M, the performance parameter time series of circuit is X
i, 1, X
i, 2, X
i, 3..., obtain the performance parameter X of institute's prediction circuit
ithe performance parameter sequence in p moment is { X continuously
i,j, X
i, j+1..., X
i, j+p, j=1,2,3 ..., p=1,2,3,
(42) utilize RVM to { X
i,j, X
i, j+1..., X
i, j+pcarry out the prediction of forward direction q step, obtain the performance parameter X of (j+p+q) moment circuit
i, j+p+q, q=1,2,3 ...
3. a kind of Power Electronic Circuit multiparameter identification of defective Forecasting Methodology based on RVM-QNN as claimed in claim 2, is characterized in that: in described step (5), the method that following certain each component parameter of moment is obtained in identification is: with X
i, j+p+qfor input, i=1,2 ..., M, picks out each component parameter θ in (j+p+q) moment
r, j+p+q, r=1,2 ..., N.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201210067840.0A CN102662142B (en) | 2012-03-15 | 2012-03-15 | Prediction method for multi-parameter identification fault of power electronic circuit based on RVM-QNN |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201210067840.0A CN102662142B (en) | 2012-03-15 | 2012-03-15 | Prediction method for multi-parameter identification fault of power electronic circuit based on RVM-QNN |
Publications (2)
Publication Number | Publication Date |
---|---|
CN102662142A CN102662142A (en) | 2012-09-12 |
CN102662142B true CN102662142B (en) | 2014-10-15 |
Family
ID=46771671
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201210067840.0A Expired - Fee Related CN102662142B (en) | 2012-03-15 | 2012-03-15 | Prediction method for multi-parameter identification fault of power electronic circuit based on RVM-QNN |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN102662142B (en) |
Families Citing this family (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104573398B (en) * | 2015-01-30 | 2017-12-26 | 安徽理工大学 | Power conversion circuit fault threshold determines method |
CN109271741B (en) * | 2018-10-25 | 2023-06-27 | 北京航空航天大学 | Method for predicting residual service life and evaluating health of step-down DC-DC power supply module |
CN111625440A (en) * | 2020-06-04 | 2020-09-04 | 中国银行股份有限公司 | Method and device for predicting performance parameters |
CN112798888B (en) * | 2020-12-30 | 2021-12-17 | 中南大学 | Non-invasive fault diagnosis method for vehicle-mounted electrical system of unmanned train |
Family Cites Families (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH01164226A (en) * | 1987-12-18 | 1989-06-28 | Fuji Electric Co Ltd | Fault diagnostic device |
-
2012
- 2012-03-15 CN CN201210067840.0A patent/CN102662142B/en not_active Expired - Fee Related
Also Published As
Publication number | Publication date |
---|---|
CN102662142A (en) | 2012-09-12 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN103886374B (en) | A kind of cable connector conductor temperature Forecasting Methodology based on RBF neural | |
CN110311376B (en) | Dynamic safety assessment comprehensive model and space-time visualization method for power system | |
Bahmanyar et al. | Power system voltage stability monitoring using artificial neural networks with a reduced set of inputs | |
CN103793853B (en) | Condition of Overhead Transmission Lines Based appraisal procedure based on two-way Bayesian network | |
CN102663412B (en) | Power equipment current-carrying fault trend prediction method based on least squares support vector machine | |
Rahi et al. | Power system voltage stability assessment through artificial neural network | |
CN102662142B (en) | Prediction method for multi-parameter identification fault of power electronic circuit based on RVM-QNN | |
Kamalasadan et al. | A new intelligent algorithm for online voltage stability assessment and monitoring | |
CN107807860B (en) | Power failure analysis method and system based on matrix decomposition | |
CN104125112A (en) | Physical-information fuzzy inference based smart power grid attack detection method | |
Yang et al. | Real-time static voltage stability assessment in large-scale power systems based on spectrum estimation of phasor measurement unit data | |
US20230118175A1 (en) | Event analysis in an electric power system | |
Zheng et al. | Real-time transient stability assessment based on deep recurrent neural network | |
Naderi et al. | Detection of false data injection cyberattacks: Experimental validation on a lab-scale microgrid | |
Zhai et al. | A degradation fault prognostic method of radar transmitter combining multivariate long short-term memory network and multivariate Gaussian distribution | |
CN111062569A (en) | Low-current fault discrimination method based on BP neural network | |
Fan et al. | Luad: A lightweight unsupervised anomaly detection scheme for multivariate time series data | |
CN103630785A (en) | Photovoltaic inverter composite particle process online residual life evaluation method | |
ul Hassan et al. | Online static security assessment for cascading failure using stacked De-noising Auto-encoder | |
Alweshah et al. | Evolution of software reliability growth models: a comparison of auto-regression and genetic programming models | |
Liang et al. | Monitoring power line faults using impedance estimation algorithms in power line communication equipment | |
Mohanta et al. | Importance and uncertainty analysis in software reliability assessment of computer relay | |
Deng et al. | Research on fault diagnosis of flexible material R2R manufacturing system based on quality control chart and SoV | |
Sampedro et al. | 3d printer state monitoring mobile application through a deep learning approach | |
CN105424224A (en) | Isolating switch state monitoring method and device |
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 | ||
CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20141015 Termination date: 20180315 |