CN112100923A - State evaluation method for frequency converter IGBT of full-power generation system - Google Patents

State evaluation method for frequency converter IGBT of full-power generation system Download PDF

Info

Publication number
CN112100923A
CN112100923A CN202010977897.9A CN202010977897A CN112100923A CN 112100923 A CN112100923 A CN 112100923A CN 202010977897 A CN202010977897 A CN 202010977897A CN 112100923 A CN112100923 A CN 112100923A
Authority
CN
China
Prior art keywords
neural network
optimized
frequency converter
full
igbt
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
Application number
CN202010977897.9A
Other languages
Chinese (zh)
Inventor
高晨
陈晓路
刘溟江
赵勇
童博
周国栋
朱亚波
韩毅
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Jiangsu Cleaning Energy Branch Of Huaneng Power Intl Inc
Huaneng Yancheng Dafeng New Energy Power Generation Co ltd
Xian Thermal Power Research Institute Co Ltd
Original Assignee
Jiangsu Cleaning Energy Branch Of Huaneng Power Intl Inc
Huaneng Yancheng Dafeng New Energy Power Generation Co ltd
Xian Thermal Power Research Institute Co Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Jiangsu Cleaning Energy Branch Of Huaneng Power Intl Inc, Huaneng Yancheng Dafeng New Energy Power Generation Co ltd, Xian Thermal Power Research Institute Co Ltd filed Critical Jiangsu Cleaning Energy Branch Of Huaneng Power Intl Inc
Priority to CN202010977897.9A priority Critical patent/CN112100923A/en
Publication of CN112100923A publication Critical patent/CN112100923A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00—Computer-aided design [CAD]
    • G06F30/20—Design optimisation, verification or simulation
    • G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
    • G01R23/00—Arrangements for measuring frequencies; Arrangements for analysing frequency spectra
    • G01R23/16—Spectrum analysis; Fourier analysis
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
    • G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
    • G01R31/50—Testing of electric apparatus, lines, cables or components for short-circuits, continuity, leakage current or incorrect line connections
    • G01R31/54—Testing for continuity
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/045—Combinations of networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/048—Activation functions
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • G06N3/084—Backpropagation, e.g. using gradient descent
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • G06N3/086—Learning methods using evolutionary algorithms, e.g. genetic algorithms or genetic programming
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/12—Computing arrangements based on biological models using genetic models
    • G06N3/126—Evolutionary algorithms, e.g. genetic algorithms or genetic programming
    • 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
    • Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
    • Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
    • Y04S10/00—Systems supporting electrical power generation, transmission or distribution
    • Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications

Abstract

The invention provides a state evaluation method of a full-power generation system frequency converter IGBT, which comprises the following steps: step 1, acquiring a 13-type waveform signal; step 2, decomposing the 13-type waveform signals obtained in the step 1 to obtain characteristic information for judging faults; step 3, constructing the characteristic information of the judgment fault obtained in the step 2 by using a wavelet decomposition coefficient to obtain a characteristic vector; step 4, extracting a characteristic value from the characteristic vector obtained in the step 3; step 5, constructing a BP neural network; step 6, optimizing the obtained BP neural network; step 7, training the optimized BP neural network in the step 6 to obtain the trained optimized BP neural network; step 8, carrying in actual operation data verification by using the optimized BP neural network trained in the step 7, and evaluating whether the IGBT of the frequency converter is open-circuited in real time; the invention solves the problem of optimal configuration in the current offshore wind power overhaul work, and improves the economy, stability and maintainability.

Description

State evaluation method for frequency converter IGBT of full-power generation system
Technical Field
The invention relates to the technical field of state evaluation of frequency converters of offshore wind power generation sets, in particular to a state evaluation method of frequency converters IGBT of a full-power generation system.
Background
The frequency converter is an important electrical device in a large-capacity offshore wind turbine generator system, is one of core energy conversion devices in the wind turbine generator system, undertakes the task of converting electric energy fluctuating along with wind speed change frequency into electric energy with stable frequency, and is also one of devices with high failure rate. At present, the mainstream offshore wind turbine generators are all full-power variable-speed constant-frequency (VSCF) power generation systems, wherein a squirrel-cage asynchronous power generation system taking a high-speed gear box, a squirrel-cage asynchronous generator and a full-power frequency converter as a technical route is common. On one hand, with the rapid development of offshore wind power industry in China, the capacity of a single machine of an offshore wind power generation unit is continuously increased, the large capacity of a converter becomes a development trend, and higher requirements are put forward on the running reliability, the economy and the maintainability of equipment; on the other hand, 12 IGBTs forming the main circuit of the two-level frequency converter are controlled to be switched on and off, the switching loss and the conduction loss of the two-level frequency converter are large, and the two-level frequency converter is a weak link of system reliability. As the offshore wind turbine generator faces more complicated environmental conditions such as high temperature, high humidity, salt spray corrosion, thunder, typhoon and the like, the wind power fluctuates violently, the electric heating stress of the wind power converter changes violently, and the security and the reliability of a main circuit power semiconductor (IGBT) of the converter are threatened. According to statistics, most faults in the main circuit of the frequency converter are caused by IGBT damage. Therefore, a new technology is needed to realize real-time state evaluation of the IGBT component of the frequency converter, discover potential fault hazards as soon as possible and provide guidance for maintenance work.
At present, no one is involved in the field of the state evaluation of the IGBT of the frequency converter of the full-power generation system.
Disclosure of Invention
The invention aims to provide a state evaluation method for a full-power generation system frequency converter IGBT, which solves the defect that the reliability of the whole system is poor because the state of the full-power generation system frequency converter IGBT is not evaluated in the prior art.
In order to achieve the purpose, the invention adopts the technical scheme that:
the invention provides a state evaluation method of a full-power generation system frequency converter IGBT, which comprises the following steps:
step 1, obtaining waveforms of a two-level full-power frequency converter in a normal working state and three-phase voltage waveform signals output by the frequency converter corresponding to the rectification circuit and the inverter circuit respectively when IGBT open circuit faults occur, and obtaining 13 types of waveform signals;
step 2, decomposing the 13-type waveform signals obtained in the step 1 to obtain characteristic information for judging faults;
step 3, constructing the characteristic information of the judgment fault obtained in the step 2 by using a wavelet decomposition coefficient to obtain a characteristic vector;
step 4, extracting a characteristic value from the characteristic vector obtained in the step 3;
step 5, constructing a BP neural network, and taking the characteristic value obtained in the step 4 as an input layer of the constructed BP neural network;
step 6, optimizing the neural network obtained in the step 5 to obtain an optimized BP neural network;
step 7, training the optimized BP neural network in the step 6 to obtain the trained optimized BP neural network;
and 8, carrying in actual operation data verification by using the optimized BP neural network trained in the step 7, and evaluating whether the IGBT of the frequency converter is open-circuited in real time.
Preferably, in step 2, the 13-type waveform signal obtained in step 1 is decomposed by a wavelet decomposition method to obtain characteristic information for determining the fault.
Preferably, in step 5, the BP neural network is constructed by the specific method:
the excitation function g (x) of the BP neural network is a Sigmoid function;
the output of the hidden layer is:
Figure BDA0002685896140000021
the output of the output layer is:
Figure BDA0002685896140000022
the mean square error is:
Figure BDA0002685896140000031
wherein, YkAs desired;
the input layer of the neural network is provided with 18 nodes; the number of the output layer neurons is determined to be 13; the number of hidden layer nodes is 20.
Preferably, the neural network obtained in step 5 is optimized, and the specific method is as follows:
and optimizing and improving the neural network by using a genetic algorithm to obtain the optimized BP neural network.
Preferably, in step 7, the BP neural network optimized in step 6 is trained, specifically:
and training the optimized BP neural network by using the training set, optimizing the weight and the threshold, and taking the optimized weight and the optimized threshold as the initial weight and the threshold of the BP neural network to obtain the trained optimized BP neural network.
Compared with the prior art, the invention has the beneficial effects that:
according to the state evaluation method of the IGBT of the frequency converter of the full-power generation system, provided by the invention, the state evaluation result of the IGBT as a key component in the two-level frequency converter can be given in real time according to the existing monitoring quantity, so that reference is provided for the maintenance of the frequency converter of the offshore full-power generation system. The problem of optimal configuration in the existing offshore wind power overhaul work can be solved, and the economy, stability and maintainability are improved;
the BP neural network has strong nonlinear mapping capability and can well establish the relation between the fault type and the fault result of the IGBT of the frequency converter; wavelet decomposition is beneficial to extracting relevant characteristic parameters. However, the traditional BP neural network also has the problems of low precision, difficulty in obtaining the optimal weight value and the like. The invention adopts the genetic algorithm improved BP neural network, and optimizes the weight and the threshold of the neural network by fusing the models established by the genetic algorithm and the neural network, thereby solving the self limitation problem of the neural network. The optimized model can well diagnose the IGBT fault of the frequency converter, the accuracy of the diagnosis result is higher, the complex frequency converter fault can be effectively positioned and diagnosed, the maintenance efficiency is improved, and the loss when the fault occurs is reduced.
Drawings
FIG. 1 is a schematic flow chart of the present invention.
Fig. 2 is a schematic diagram of a two-level PWM-type frequency converter.
Fig. 3 is a flow chart of the improved BP neural network.
Fig. 4 is an improved BP neural network training diagram.
FIG. 5 is a diagram of an improved BP neural network prediction.
Detailed Description
The present invention will be described in further detail with reference to the accompanying drawings.
As shown in fig. 1, the method for evaluating the state of the inverter IGBT of the full power generation system according to the present invention includes the following steps:
step 1, collecting and screening three-phase voltage waveform data of a network side in the running process of a two-level full-power frequency converter, and respectively obtaining waveforms of the two-level full-power frequency converter in a normal working state and three-phase voltage waveform signals output by respectively corresponding frequency converters when IGBTs of a rectifying circuit and an inverter circuit are in open-circuit faults, wherein 13 types of waveform signals are respectively corresponding to a normal state or a fault state;
step 2, decomposing the 13-type waveform signals by using wavelet decomposition to obtain characteristic information for judging faults;
step 3, constructing the characteristic information of the judgment fault obtained in the step 2 by using a wavelet decomposition coefficient to obtain a characteristic vector;
step 4, extracting a normalized characteristic value from the obtained characteristic vector, taking the characteristic value as a neural network input layer, and entering step 5;
step 5, setting a network hidden layer and an output layer, constructing a BP neural network, and training the constructed BP neural network to obtain a trained BP neural network;
step 6, utilizing a multi-population algorithm to improve and optimize the trained BP neural network to obtain an optimized BP neural network;
step 7, training the optimized BP neural network to obtain the trained optimized BP neural network;
and 8, carrying in actual operation data verification by using the trained optimized BP neural network, and evaluating whether the IGBT of the frequency converter is open-circuited in real time.
Specifically, the method comprises the following steps:
the method for decomposing the 13 types of waveform signals by using wavelet decomposition comprises the following specific steps:
setting wavelet decomposition as 5-layer decomposition form, and decomposing the original signal as follows:
X＝A5+D1+D2+D3+D4+D5
wherein, X is an original waveform signal; a. the5Decomposing the high-frequency part contained in the fifth layer; d1、D2、D3、D4And D5The low-frequency part contained after each layer is decomposed;
A5、D1、D2、D3、D4and D5Corresponding wavelet packet decomposition coefficients are arranged on all the parts, and feature vectors are constructed according to the coefficients obtained by decomposition:
SA＝[C1 C2 C3 C4 C5 C6]
C1、C2、C3、C4、C5and C6Is the decomposition coefficient of each layer of phase A, SAIs A-phase eigenvector;
S＝[SA SB SC]
wherein S is a three-phase eigenvector;
normalizing the elements in the feature vector S by the following formula to obtain a normalized feature value:
Figure BDA0002685896140000051
wherein y is a normalized value, x is a sample value, and x isminIs the smallest value, x, of the values in the sample datamaxIs the maximum value of the data in the sample.
The experiment of each group actually extracts a feature vector consisting of 18 feature elements from the waveform of each fault.
In the step 5, a BP neural network algorithm is adopted to construct a model in the step 6, and the specific method is as follows:
algorithm design
Selecting the characteristic value obtained in the step 4, firstly setting an excitation function g (x) as a Sigmoid function
Figure BDA0002685896140000061
Setting input layer to hidden layer weight value (omega)ij) Threshold value (a)j) Hidden layer to output layer weight (ω)jk) Threshold value (b)k) The number of nodes of the input layer, the hidden layer and the output layer is n, l, m, and the hidden layer output (H) can be obtainedj) And output layer output (O)k) Can be expressed as:
Figure BDA0002685896140000062
Figure BDA0002685896140000063
mean square error (E) of output layerk) Can be expressed as:
Figure BDA0002685896140000064
Ykas desired.
And observing whether the error is in an allowable range according to the output, if not, continuing the operation, and if so, ending the operation.
Designing an input layer, wherein the fault characteristic vectors obtained by wavelet analysis can be known, the input layer of the neural network is designed with 18 nodes, and the sample fault data is 18-dimensional;
and in the design of an output layer, aiming at the single-tube open-circuit fault of the frequency converter, 13 fault states are provided, namely the state when the frequency converter operates normally and the state when 12 IGBTs have open-circuit faults respectively, so that the number of neurons in the output layer is determined to be 13. Failure modes are defined herein as 1, 2, 3, …, 13. Wherein 1 represents the normal state of the frequency converter, and 2 to 13 represent the fault states corresponding to the IGBTs with different numbers.
Adopting binary fault codes of 13 bits for the output of the neural network and respectively corresponding to different fault modes;
and (4) designing the hidden layer, obtaining the number of nodes of the hidden layer by using an empirical formula, and comparing the number of the nodes of the hidden layer to select an optimal solution. The number of the hidden layer nodes is finally determined to be 20 through repeated tests.
Training function setting, wherein the training function of the neural network selects train lm, the transfer function from the input layer to the hidden layer selects a distance function, and the function from the hidden layer to the output layer selects a purelin function.
Preferably, the BP neural network algorithm is improved by using a genetic algorithm in step 6:
the Genetic Algorithm (Genetic Algorithm) can improve the global search capability of the neural network and can well solve the problem that the neural network is easy to fall into local minimum. A multi-population algorithm is utilized herein to improve the BP neural network.
Optimization and improvement by using a genetic algorithm, specifically:
firstly, coding the weight and the threshold of the BP neural network;
and selecting a fitness function, wherein the fitness function is a standard for distinguishing the individual from good or bad, and is a function for calculating the fitness of the individual determined according to the evolution target. The functional formula is expressed as:
Figure BDA0002685896140000071
n is the number of output nodes of the neural network, riIs the actual value of the ith node in the neural network, piK is a coefficient, which is a predicted value of the ith node.
Selecting operation is carried out by using a roulette algorithm, and population evolution is carried out by using a real number intersection method:
Figure BDA0002685896140000072
setting a variation function:
Figure BDA0002685896140000073
amaxand aminIs aijF (g) is the evolution function, g is the current evolution times.
And taking the optimized weight threshold value as an initial weight and a threshold value of the neural network.
The network is trained using a training set.
In step 8, the specific method for evaluating the frequency converter IGBT according to the established improved BP neural network is as follows:
collecting three-phase voltage data of a frequency converter network side, and extracting a characteristic vector by utilizing 5-layer wavelet decomposition;
optimizing the BP neural network by using a genetic algorithm, wherein the parameters are set as follows:
TABLE 1 genetic Algorithm parameter set-ups
Figure BDA0002685896140000081
And taking the optimized weight threshold value as an initial weight and a threshold value of the neural network, and operating the neural network.
And 8, evaluating the actual states of the 12 IGBTs in the frequency converter according to the 13-dimensional vector output in the step 7.
Examples
The full-power generation system adopts a two-level mode, a core rectification inverter circuit is composed of 12 high-power IGBTs, and the whole structure is shown in figure 2; specifically, the method comprises the following steps:
collecting three-phase voltage operation data of a frequency converter network side;
the data is extracted by wavelet decomposition, and the number of decomposition layers is 5.
X＝A5+D1+D2+D3+D4+D5
Wherein, X is an original waveform signal; a. the5Decomposing the high-frequency part contained in the fifth layer; d1、D2、D3、D4And D5The low frequency part contained after the decomposition of each layer.
Extracting each decomposition coefficient to construct a feature vector:
SA＝[C1 C2 C3 C4 C5 C6]
C1、C2、C3、C4、C5and C6Is a decomposition coefficient of each layer, SAIs A-phase eigenvector;
S＝[SA SB SC]
s is a three-phase eigenvector;
and (3) establishing a BP neural network by taking a genetic algorithm as an optimization strategy, as shown in figure 3.
Normalizing elements in the feature vector as follows:
Figure BDA0002685896140000091
wherein y is a normalized value, x is a sample value, and x isminIs the smallest value, x, of the values in the sample datamaxIs the maximum value of the data in the sample.
The excitation function is set to be a Sigmoid function,
Figure BDA0002685896140000092
setting the weight value of the hidden layer, the threshold value and the learning rate of the input layer, the output of the hidden layer and the output layer can be expressed as:
Figure BDA0002685896140000093
Figure BDA0002685896140000094
the output mean square error can be expressed as:
Figure BDA0002685896140000095
coding the weight and the threshold of the BP neural network;
and selecting a fitness function, wherein the fitness function is a standard for distinguishing the individual from good or bad, and is a function for calculating the fitness of the individual determined according to the evolution target. The functional formula is expressed as:
Figure BDA0002685896140000101
n is the number of output nodes of the neural network, riIs the actual value of the ith node in the neural network, piK is a coefficient, which is a predicted value of the ith node.
Selecting operation is carried out by using a roulette algorithm, and population evolution is carried out by using a real number intersection method:
Figure BDA0002685896140000102
setting a variation function:
Figure BDA0002685896140000103
amaxand aminIs aijF (g) is the evolution function, g is the current evolution times.
And taking the optimized weight threshold value as an initial weight and a threshold value of the neural network, and training the network by using a training set. The training results are shown in fig. 4.
And continuously adjusting the evolution function, and finally optimizing the weight range. For the squirrel cage full power generation system. The state of each IGBT is predicted by adopting the improved BP neural network, and the accuracy rate can reach more than 95 percent, as shown in figure 5. The operation and maintenance analysis efficiency and accuracy are greatly improved.

Claims (5)

1. A state evaluation method for a frequency converter IGBT of a full-power generation system is characterized by comprising the following steps:
step 1, obtaining waveforms of a two-level full-power frequency converter in a normal working state and three-phase voltage waveform signals output by the frequency converter corresponding to the rectification circuit and the inverter circuit respectively when IGBT open circuit faults occur, and obtaining 13 types of waveform signals;
step 2, decomposing the 13-type waveform signals obtained in the step 1 to obtain characteristic information for judging faults;
step 3, constructing the characteristic information of the judgment fault obtained in the step 2 by using a wavelet decomposition coefficient to obtain a characteristic vector;
step 4, extracting a characteristic value from the characteristic vector obtained in the step 3;
step 5, constructing a BP neural network, and taking the characteristic value obtained in the step 4 as an input layer of the constructed BP neural network;
step 6, optimizing the neural network obtained in the step 5 to obtain an optimized BP neural network;
step 7, training the optimized BP neural network in the step 6 to obtain the trained optimized BP neural network;
and 8, carrying in actual operation data verification by using the optimized BP neural network trained in the step 7, and evaluating whether the IGBT of the frequency converter is open-circuited in real time.
2. The method for evaluating the state of the inverter IGBT of the full-power generation system according to claim 1, characterized in that in step 2, the 13-type waveform signals obtained in step 1 are decomposed by a wavelet decomposition method to obtain characteristic information for judging faults.
3. The state evaluation method of the full-power generation system frequency converter IGBT according to claim 1, characterized in that in step 5, a BP neural network is constructed, and the specific method is as follows:
the excitation function g (x) of the BP neural network is a Sigmoid function;
the output of the hidden layer is:
Figure FDA0002685896130000011
the output of the output layer is:
Figure FDA0002685896130000021
the mean square error is:
Figure FDA0002685896130000022
wherein, YkAs desired;
the input layer of the neural network is provided with 18 nodes; the number of the output layer neurons is determined to be 13; the number of hidden layer nodes is 20.
4. The state evaluation method of the full-power generation system frequency converter IGBT according to claim 1, characterized in that the neural network obtained in step 5 is optimized, and the specific method is as follows:
and optimizing and improving the neural network by using a genetic algorithm to obtain the optimized BP neural network.
5. The state evaluation method of the full-power generation system frequency converter IGBT according to claim 1, characterized in that in step 7, the BP neural network optimized in step 6 is trained, specifically:
and training the optimized BP neural network by using the training set, optimizing the weight and the threshold, and taking the optimized weight and the optimized threshold as the initial weight and the threshold of the BP neural network to obtain the trained optimized BP neural network.
CN202010977897.9A 2020-09-16 2020-09-16 State evaluation method for frequency converter IGBT of full-power generation system Pending CN112100923A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202010977897.9A CN112100923A (en) 2020-09-16 2020-09-16 State evaluation method for frequency converter IGBT of full-power generation system

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202010977897.9A CN112100923A (en) 2020-09-16 2020-09-16 State evaluation method for frequency converter IGBT of full-power generation system

Publications (1)

Publication Number Publication Date
CN112100923A true CN112100923A (en) 2020-12-18

Family

ID=73758799

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202010977897.9A Pending CN112100923A (en) 2020-09-16 2020-09-16 State evaluation method for frequency converter IGBT of full-power generation system

Country Status (1)

Country Link
CN (1) CN112100923A (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112836738A (en) * 2021-01-29 2021-05-25 华能国际电力股份有限公司 Electricity stealing behavior detection method based on BP neural network
CN114964370A (en) * 2022-05-25 2022-08-30 国家电投集团科学技术研究院有限公司 Wind turbine generator set frequency converter state monitoring method and system and electronic equipment

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20060266044A1 (en) * 2005-04-25 2006-11-30 Frank Donnelly Alternator boost method
CN101017980A (en) * 2007-02-15 2007-08-15 湖南大学 Multi-variant control method of the distribution static reactive power generator
US20130096858A1 (en) * 2011-10-17 2013-04-18 International Business Machines Corporation System, method, and program for predicting state of battery
CN104401036A (en) * 2014-10-22 2015-03-11 宁波步络科工业自动化科技有限公司 Brake curve self-learning method of numerical-control punch press based on BP neural network

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20060266044A1 (en) * 2005-04-25 2006-11-30 Frank Donnelly Alternator boost method
CN101017980A (en) * 2007-02-15 2007-08-15 湖南大学 Multi-variant control method of the distribution static reactive power generator
US20130096858A1 (en) * 2011-10-17 2013-04-18 International Business Machines Corporation System, method, and program for predicting state of battery
CN104401036A (en) * 2014-10-22 2015-03-11 宁波步络科工业自动化科技有限公司 Brake curve self-learning method of numerical-control punch press based on BP neural network

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
赵海龙: "基于Matlab故障仿真的变频器主电路故障诊断方法研究", 《中国优秀硕士学位论文全文数据库 工程科技Ⅱ辑》 *

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112836738A (en) * 2021-01-29 2021-05-25 华能国际电力股份有限公司 Electricity stealing behavior detection method based on BP neural network
CN112836738B (en) * 2021-01-29 2023-05-05 华能国际电力股份有限公司 BP neural network-based electricity stealing behavior detection method
CN114964370A (en) * 2022-05-25 2022-08-30 国家电投集团科学技术研究院有限公司 Wind turbine generator set frequency converter state monitoring method and system and electronic equipment

Similar Documents

Publication Publication Date Title
CN106443297B (en) The decision tree SVM method for diagnosing faults of photovoltaic diode Clamp three-level inverter
CN108832619A (en) Transient stability evaluation in power system method based on convolutional neural networks
CN110619321B (en) Charging pile technical condition diagnosis method and system based on deep neural network
CN111144638B (en) Power distribution network operation situation prediction method based on big data
CN104408562A (en) Photovoltaic system generating efficiency comprehensive evaluation method based on BP (back propagation) neural network
CN112100923A (en) State evaluation method for frequency converter IGBT of full-power generation system
CN111007429A (en) ANFIS-based synchronous phase modulator short-circuit fault identification method and system
Geng et al. Fault diagnosis of converter based on wavelet decomposition and BP neural network
CN116505656A (en) Wind-light-storage multifunctional complementary intelligent power utilization system based on 5G Internet of things technology
CN111062569A (en) Low-current fault discrimination method based on BP neural network
An et al. IGBT open circuit fault diagnosis method for a modular multilevel converter based on PNN-MD
CN115469184A (en) New energy transmission line fault identification method based on convolutional network
CN115037054A (en) Commutation switch intelligent terminal based on three-phase self-balancing and low-voltage fault location isolation
Mu et al. Cascading Fault Early Warning and Location Method of Transmission Networks Based on Wide Area Time-Series Power System State
CN114462692A (en) Power grid old and old equipment technical improvement strategy optimization and adjustment method
CN113725910A (en) Stability analysis and quantitative evaluation method for wind power plant grid-connected system
Hui et al. The application of artificial intelligence technology in the fault diagnosis of floating wind turbine generator
CN110957723A (en) Data-driven method for rapidly evaluating transient voltage safety of power grid on line
Huang et al. Research of Transformer Protection Based on Joint Deep Learning
CN118013343A (en) Electrical equipment fault prediction method based on CSO-ANN algorithm
CN113991652B (en) Data-driven multi-output calculation method for short-circuit current of distribution network containing IIDG
Ruirong et al. Research on Fault Location Technology of Intelligent Distribution Network based on Neural Network
Zhang et al. A Novel Data-Driven Fault Diagnosis Method Based on Semi-Generative Adversarial Network for Buck Converter ICs
CN115034493B (en) Wind farm black start path optimizing method considering unit running state
Wang et al. Fault diagnosis method for hybrid HVDC transmission system based on stacked sparse auto-encoder

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: 20201218

RJ01 Rejection of invention patent application after publication