CN111881628A - Method and system for optimizing electrical parameters of square wave voltage source and storage medium - Google Patents

Method and system for optimizing electrical parameters of square wave voltage source and storage medium Download PDF

Info

Publication number
CN111881628A
CN111881628A CN202010784036.9A CN202010784036A CN111881628A CN 111881628 A CN111881628 A CN 111881628A CN 202010784036 A CN202010784036 A CN 202010784036A CN 111881628 A CN111881628 A CN 111881628A
Authority
CN
China
Prior art keywords
optimization
square wave
voltage source
wave voltage
electrical
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN202010784036.9A
Other languages
Chinese (zh)
Other versions
CN111881628B (en
Inventor
翟少磊
魏龄
张林山
沈鑫
王恩
邓涛
陈文华
刘静
李登云
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.)
Electric Power Research Institute of Yunnan Power Grid Co Ltd
Original Assignee
Electric Power Research Institute of Yunnan Power Grid 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 Electric Power Research Institute of Yunnan Power Grid Co Ltd filed Critical Electric Power Research Institute of Yunnan Power Grid Co Ltd
Priority to CN202010784036.9A priority Critical patent/CN111881628B/en
Publication of CN111881628A publication Critical patent/CN111881628A/en
Application granted granted Critical
Publication of CN111881628B publication Critical patent/CN111881628B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

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
    • 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
    • G06N3/08—Learning methods
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2111/00—Details relating to CAD techniques
    • G06F2111/10—Numerical modelling
    • 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
    • Y02E—REDUCTION OF GREENHOUSE GAS [GHG] EMISSIONS, RELATED TO ENERGY GENERATION, TRANSMISSION OR DISTRIBUTION
    • Y02E60/00—Enabling technologies; Technologies with a potential or indirect contribution to GHG emissions mitigation
    • Y02E60/60—Arrangements for transfer of electric power between AC networks or generators via a high voltage DC link [HVCD]

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Evolutionary Computation (AREA)
  • Software Systems (AREA)
  • Artificial Intelligence (AREA)
  • General Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • Computing Systems (AREA)
  • Mathematical Physics (AREA)
  • General Health & Medical Sciences (AREA)
  • Molecular Biology (AREA)
  • Computational Linguistics (AREA)
  • Biophysics (AREA)
  • Biomedical Technology (AREA)
  • Data Mining & Analysis (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Health & Medical Sciences (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Medical Informatics (AREA)
  • Computer Hardware Design (AREA)
  • Geometry (AREA)
  • Supply And Distribution Of Alternating Current (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

The invention discloses a method, a system and a storage medium for square wave voltage source electrical parameter optimization, wherein the method for square wave voltage source electrical parameter optimization comprises the following steps: constructing a mathematical model of a square wave voltage source voltage forming loop, determining electrical optimization parameters and optimization targets of the square wave voltage source voltage forming loop, and determining a value range of the electrical optimization parameters; and calculating to obtain a training sample set for training the network through the value range of the electrical optimization parameters, then training through the deep belief network to obtain the trained deep belief network, and finally obtaining a group of parameters with the minimum comprehensive optimization target, namely the electrical parameter optimal value. According to the square wave voltage source electrical parameter optimization method, the deep belief network is used, so that the test times and the test time are reduced, the deep belief network is trained firstly, namely the deep belief network is pre-trained, and the initial parameters of the network are more accurate.

Description

Method and system for optimizing electrical parameters of square wave voltage source and storage medium
Technical Field
The invention relates to the field of high-voltage direct-current transmission, in particular to a method and a system for optimizing electrical parameters of a square wave voltage source and a storage medium.
Background
The direct current transformer is used as an important primary device in a high-voltage direct current transmission system, and the accuracy, reliability and transient characteristics of the direct current transformer directly influence the safe and stable operation of the high-voltage direct current transmission system. Particularly, with the wide application of the flexible direct current system, the control protection device of the flexible direct current system puts higher requirements on the rapidity and the accuracy of the transient response of the direct current transformer. However, due to the lack of technical equipment for correspondingly checking and calibrating transient characteristics, the direct current transformer is not reliably and effectively checked and calibrated all the time. According to actual operation and test results of direct current transformers in the current domestic converter stations, various direct current transformer devices have certain problems and hidden dangers, partial direct current transformers of the converter stations have faults for many times, and some faults directly cause monopole locking of a direct current system, so that great threats and hidden dangers are brought to stable and reliable power grid and economic operation. The relevant test results show that some performance defects can be found early through field calibration or periodic calibration, and some performance defects can be prevented and corrected through field tests. Therefore, the square wave voltage source for field test of the direct current measuring device is required to perform transient characteristic verification and calibration on the high-voltage direct current transformer, so that the safe operation of the direct current transmission system is ensured.
At present, methods for performing transient characteristic verification and calibration on a high-voltage direct-current transformer by applying a square-wave voltage source are an exhaustion method and a neural network method, and the exhaustion method requires many test times, so that the test time is long; neural network methods do not have pre-training and therefore do not have better initial parameters for the network.
Therefore, how to design a method for optimizing the electrical parameters of the square wave voltage source with less test times and accurate initial parameters of the network becomes an urgent problem to be solved.
Disclosure of Invention
The invention provides a method, a system and a storage medium for square wave voltage source electrical parameter optimization, which aim to solve the problems of multiple test times and inaccurate network initial parameters of the existing square wave voltage source electrical parameter optimization method.
In a first aspect, the present invention provides a method for electrical parameter optimization of a square wave voltage source, comprising:
s1: constructing a mathematical model of a square wave voltage source voltage forming loop, and solving through MATLAB to obtain a square wave voltage source state expression, wherein the square wave voltage source state expression comprises a direct current voltage source voltage expression, a load voltage expression, an energy storage capacitor expression and a load capacitor expression;
s2: determining electrical optimization parameters and optimization targets of a square wave voltage source voltage forming loop according to a square wave voltage source state expression, and determining the value range of the electrical optimization parameters by combining a mapping method and a load voltage expression, wherein the electrical optimization parameters comprise an energy storage capacitor, a current limiting resistor and a pull-down resistor, the optimization targets comprise the rise time, the duration and the maximum current of the square wave voltage, and the value range of the electrical optimization parameters comprises the value range of the energy storage capacitor, the value range of the current limiting resistor and the value range of the pull-down resistor;
s3: calculating to obtain a training sample set for training the network according to the value range of the electrical optimization parameters;
s4: training through a deep belief network according to a training sample set, and calculating the relation between each electrical parameter of a square wave voltage source voltage forming loop and an optimization target until the deep belief network meets the error requirement to obtain a trained deep belief network;
s5: unifying the three optimization targets into a comprehensive optimization target;
s6: in the trained deep belief network, a group of parameters with the minimum comprehensive optimization target is obtained through the parameters, and the group of parameters with the minimum comprehensive optimization target is the optimal value of the electrical parameter.
Optionally, the calculating, through the value range of the electrical optimization parameter, a training sample set for training the network includes:
s31: selecting an input sample within the value range of the electrical optimization parameter;
s32: output samples under input samples, namely a training sample set for training the network, are obtained through MATLAB/SIMULINK simulation.
Optionally, each of the input samples includes an energy storage capacitance value, a current limiting resistance value and a pull-down resistance value.
Optionally, the three optimization targets are unified into one comprehensive optimization target, and the weight of the optimization target is set by an optimization target subjective method and an entropy weight method to obtain the comprehensive optimization target.
In a second aspect, the present invention provides a system for optimizing electrical parameters of a square wave voltage source, including a square wave voltage source state expression obtaining module, an electrical optimization parameter-optimization target obtaining module, a training sample set obtaining module, a training module, a comprehensive optimization target obtaining module, and an electrical parameter optimal value obtaining module, wherein:
the square wave voltage source state expression obtaining module is used for constructing a mathematical model of a square wave voltage source voltage forming loop, and solving through MATLAB to obtain a square wave voltage source state expression, wherein the square wave voltage source state expression comprises a direct current voltage source voltage expression, a load voltage expression, an energy storage capacitor expression and a load capacitor expression;
the electrical optimization parameter-optimization target acquisition module is used for determining electrical optimization parameters and optimization targets of a square wave voltage source voltage forming loop according to a square wave voltage source state expression, combining a mapping method and a load voltage expression to determine a value range of the electrical optimization parameters, wherein the electrical optimization parameters comprise an energy storage capacitor, a current limiting resistor and a pull-down resistor, the optimization targets comprise square wave voltage rise time, duration and maximum current, and the value range of the electrical optimization parameters comprises the value range of the energy storage capacitor, the value range of the current limiting resistor and the value range of the pull-down resistor;
the training sample set acquisition module is used for calculating a training sample set used for training the network according to the value range of the electrical optimization parameters;
the training module is used for training through a deep belief network according to a training sample set, calculating the relation between each electrical parameter of a square wave voltage source voltage forming loop and an optimization target until the deep belief network meets the error requirement, and obtaining the trained deep belief network;
the comprehensive optimization target obtaining module is used for setting the weight of the optimization target through a subjective method and an entropy weight method, and unifying the three optimization targets into one comprehensive optimization target;
the electrical parameter optimal value acquisition module is used for acquiring a group of parameters with the minimum comprehensive optimization target through parameters in a trained deep belief network, wherein the group of parameters with the minimum comprehensive optimization target is the electrical parameter optimal value.
In a third aspect, the present invention provides a storage medium containing computer executable instructions which, when executed by a computer processor, implement the method of square wave voltage source electrical parameter optimization as described in the first aspect.
The invention provides a method, a system and a storage medium for square wave voltage source electrical parameter optimization, wherein the method for square wave voltage source electrical parameter optimization comprises the following steps: constructing a mathematical model of a square wave voltage source voltage forming loop, and solving through MATLAB to obtain a square wave voltage source state expression; determining electrical optimization parameters and optimization targets of a square wave voltage source voltage forming loop according to a square wave voltage source state expression, and determining a value range of the electrical optimization parameters by combining a mapping method and a load voltage expression; calculating to obtain a training sample set for training the network according to the value range of the electrical optimization parameters; training through a deep belief network according to a training sample set, and calculating the relation between each electrical parameter of a square wave voltage source voltage forming loop and an optimization target until the deep belief network meets the error requirement to obtain a trained deep belief network; unifying the optimization targets into a comprehensive optimization target; in the trained deep belief network, a group of parameters with the minimum comprehensive optimization target is obtained through the parameters, and the group of parameters with the minimum comprehensive optimization target is the optimal value of the electrical parameter. According to the square wave voltage source electrical parameter optimization method, the deep belief network is used, so that the test times and the test time are reduced, the deep belief network is trained firstly, namely the deep belief network is pre-trained, and the initial parameters of the network are more accurate.
Drawings
In order to more clearly illustrate the technical solution of the present invention, the drawings needed to be used in the embodiments will be briefly described below, and it is obvious to those skilled in the art that other drawings can be obtained according to the drawings without any inventive exercise.
FIG. 1 is a flow chart of a method for electrical parameter optimization of a square wave voltage source according to the present invention;
FIG. 2 is a deep belief network training result of the method for optimizing electrical parameters of a square wave voltage source provided by the present invention;
FIG. 3 is a schematic diagram of a voltage forming loop for a square wave voltage source according to the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention are clearly and completely described below with reference to the drawings in the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
In the following description, numerous specific details are set forth in order to provide a thorough understanding of the present invention, but the present invention may be practiced in other ways than those specifically described, and it will be appreciated by those skilled in the art that the present invention may be embodied without departing from the spirit and scope of the invention, and therefore the present invention is not limited to the specific embodiments disclosed below.
Referring to fig. 1, in a first aspect, the present invention provides a method for electrical parameter optimization of a square wave voltage source, comprising:
s1: constructing a mathematical model of a square wave voltage source voltage forming loop, and solving through MATLAB to obtain a square wave voltage source state expression, wherein the square wave voltage source state expression comprises a direct current voltage source voltage expression, a load voltage expression, an energy storage capacitor expression and a load capacitor expression;
the square wave voltage source state expression is as follows:
Figure BDA0002621267090000041
in the formula: u1 is direct-current voltage source voltage, u2 is load voltage, R1 and R2 are respectively a current-limiting resistor and a pull-down resistor, and C1 and C2 are respectively an energy storage capacitor and a load capacitor.
S2: determining electrical optimization parameters and optimization targets of a square wave voltage source voltage forming loop according to a square wave voltage source state expression, and determining the value range of the electrical optimization parameters by combining a mapping method and a load voltage expression, wherein the electrical optimization parameters comprise an energy storage capacitor, a current limiting resistor and a pull-down resistor, the optimization targets comprise the rise time, the duration and the maximum current of the square wave voltage, and the value range of the electrical optimization parameters comprises the value range of the energy storage capacitor, the value range of the current limiting resistor and the value range of the pull-down resistor;
the electrical optimization parameters of the square wave voltage forming loop are as follows:
C1,R1,R2
c1 is the energy storage capacitor of square wave voltage source with the value range of [ C1min,C1max]；
R1 is the current-limiting resistance of square wave voltage source, and the value range is [ R1min,R1max]；
R2 is pull-down resistor of square wave voltage source with value range of [ R2min,R2max]；
The optimization target is as follows: tr, ts, imax
tr is the rise time of the square wave voltage source, and the smaller the value is, the better the value is;
ts is a voltage holding time, and the larger the value, the better;
imax is the peak value of the loop current, and the smaller the value, the better;
the electrical optimization value range is shown in table 1.
Structural parameters Range of variation
C1/μF [0.3,1]
R1/Ω [1500,3500]
R2/MΩ [4,20]
TABLE 1 electric optimization parameter variation Range interval
The experimental factors included: the capacitor comprises a storage capacitance value C1, a current limiting resistance value R1 and a pull-down resistance value R2.
S3: calculating to obtain a training sample set for training the network according to the value range of the electrical optimization parameters;
s4: training through a deep belief network according to a training sample set, and calculating the relation between each electrical parameter of a square wave voltage source voltage forming loop and an optimization target until the deep belief network meets the error requirement to obtain a trained deep belief network;
s5: unifying the three optimization targets into a comprehensive optimization target;
s6: in the trained deep belief network, a group of parameters with the minimum comprehensive optimization target is obtained through the parameters, and the group of parameters with the minimum comprehensive optimization target is the optimal value of the electrical parameter.
Optionally, the calculating, through the value range of the electrical optimization parameter, a training sample set for training the network includes:
s31: selecting an input sample within the value range of the electrical optimization parameter;
s32: output samples under input samples, namely a training sample set for training the network, are obtained through MATLAB/SIMULINK simulation.
Optionally, each of the input samples includes an energy storage capacitance value, a current limiting resistance value and a pull-down resistance value.
Selecting a plurality of training samples in the value range of the electrical optimization parameters, and obtaining square wave voltage technical index values (including tr, ts and imax) corresponding to each group of electrical parameters by using MATLAB/SIMULINK simulation to construct a training sample set TRAIN (p 3);
the selection mode of the training sample is as follows: in [ C ]1min,C1max]Taking p points as energy storage capacitance parameters at [ R ] by inner average1min,R1max]Taking p points as current limiting resistance parameters in inner average at R2min,R2max]Taking p points as pull-down resistance parameters by inner averaging;
then, the p is selected out by permutation and combination3In a combinatorial manner, construct p3Each training sample constituting a training sample set TRAIN (p)3)；
Each training sample comprises an energy storage capacitance value, a current limiting resistance value and a pull-down resistance value, and technical indexes such as corresponding square wave voltage rise time tr, voltage holding time ts, loop current peak value imax and the like are obtained according to MATLAB/SIMULINK simulation.
The deep belief network training comprises the following specific steps:
step 1, pre-training stage: the pre-training stage is a process of training a multi-layer Restricted Boltzmann Machine (RBM), and the pre-training starts from a bottom RBM and is sequentially trained from bottom to top.
The RBM is provided with a visible layer V and a hidden layer H; at a given kth RBM model parameter thetak＝{Wk,ak,bk-defining an energy function for this RBM:
Figure BDA0002621267090000051
wherein, Vk＝[vk,1,vk,2,...,vk,m]TIs the state vector of the visible layer of the kth RBM unit, vk,iRepresents the state of the ith neuron of the kth RBM unit, i ∈ [1, m [ ]]M represents the number of neurons in the visible layer, k ∈ [1, L ]]L represents the number of RBM units;
Ak＝[ak,1,ak,2,...,ak,m]Tis the bias vector of the visible layer of the k-th RBM unit, ak,iDenotes the bias of the ith neuron of the kth RBM Unit, i ∈ [1, m [ ]]L represents the number of RBM units;
Hk＝[hk,1,hk,2,...,hk,n]Tis the state vector of the hidden layer of the kth RBM unit, hk,jRepresents the state of the jth neuron of the kth RBM unit, j ∈ [1, n ]]N represents the number of hidden layer neurons, and L represents the number of RBM units;
Bk＝[bk,1,bk,2,...,bk,n]Tis the bias vector of the hidden layer of the k-th RBM unit, bk,jRepresents the bias of the jth neuron of the kth RBM unit, j ∈ [1, n ]]L represents the number of RBM units;
Wk＝{wk,i,j}∈RL×m×na weight matrix representing the kth RBM unit connecting neurons of a visible layer and neurons of a hidden layer;
defining a function
Figure BDA0002621267090000061
The index is used for judging the kth RBM training result, the larger the value of the index is, the more fit between the RBM pre-training result and the distribution of training samples is represented, and the correlation process is described by the following mathematical formula:
Figure BDA0002621267090000062
Figure BDA0002621267090000063
in the above formula, θk＝{Wk,ak,bkS is a training sample set, nsIn order to train the number of samples,
Figure BDA0002621267090000064
each sample in the set S has the same probability distribution and is independent of each other;
step 2, fine adjustment stage: after the RBM pre-training is finished, the initial parameter theta after the training is finishedk＝L＝{Wk＝L,ak＝L,bk＝LAssigning to the last layer of neural network; and optimizing the whole network from top to bottom by adopting a Levenberg-Marquardt algorithm and taking a training sample as a supervision signal.
The evaluation index of the deep belief network training is the average absolute percentage error eMAPE，eMAPEIs expressed as
Figure BDA0002621267090000065
In the formula, yiAnd
Figure BDA0002621267090000066
outputting the actual value and the predicted value of the optimization target for the predicted point, eMAPEThe smaller the value, the more accurate the result prediction.
And training by utilizing Matlab software and a deep belief network method according to the obtained training samples. Wherein, 3 layers of RBMs are selected for pre-training, and the number of hidden layer units is 20, 10 and 5 from bottom to top in sequence. And assigning the initial parameters after training to the last layer of BP neural network, and training by using an LM (Levenberg-Marquardt) algorithm. And fitting the relation between each structural parameter of the grading ring and the optimization target to obtain a trained network.
Setting the weight of the optimization target by using a subjective method and an entropy weight method, unifying the three optimization targets into a comprehensive optimization target, wherein the weights of tr, ts and imas are respectively 0.2521, 0.4243 and 0.3244; in the trained deep belief network, a group of parameters with the minimum comprehensive optimization target is selected from the parameters throughout the process, and the parameters are the optimal values of the electrical parameters.
Optionally, the three optimization targets are unified into one comprehensive optimization target, and the weight of the optimization target is set by an optimization target subjective method and an entropy weight method to obtain the comprehensive optimization target.
Referring to fig. 2 and 3, the best scheme of the grading ring structure obtained by using the trained network is as follows: c1 ═ 0.42 μ F, R1 ═ 1.8k Ω, and R2 ═ 5.3M, where the square-wave voltage index tr is 11.4 μ s, ts is 21ms, and imax is 51A.
In a second aspect, the present invention provides a system for optimizing electrical parameters of a square wave voltage source, including a square wave voltage source state expression obtaining module, an electrical optimization parameter-optimization target obtaining module, a training sample set obtaining module, a training module, a comprehensive optimization target obtaining module, and an electrical parameter optimal value obtaining module, wherein:
the square wave voltage source state expression obtaining module is used for constructing a mathematical model of a square wave voltage source voltage forming loop, and solving through MATLAB to obtain a square wave voltage source state expression, wherein the square wave voltage source state expression comprises a direct current voltage source voltage expression, a load voltage expression, an energy storage capacitor expression and a load capacitor expression;
the electrical optimization parameter-optimization target acquisition module is used for determining electrical optimization parameters and optimization targets of a square wave voltage source voltage forming loop according to a square wave voltage source state expression, combining a mapping method and a load voltage expression to determine a value range of the electrical optimization parameters, wherein the electrical optimization parameters comprise an energy storage capacitor, a current limiting resistor and a pull-down resistor, the optimization targets comprise square wave voltage rise time, duration and maximum current, and the value range of the electrical optimization parameters comprises the value range of the energy storage capacitor, the value range of the current limiting resistor and the value range of the pull-down resistor;
the training sample set acquisition module is used for calculating a training sample set used for training the network according to the value range of the electrical optimization parameters;
the training module is used for training through a deep belief network according to a training sample set, calculating the relation between each electrical parameter of a square wave voltage source voltage forming loop and an optimization target until the deep belief network meets the error requirement, and obtaining the trained deep belief network;
the comprehensive optimization target obtaining module is used for setting the weight of the optimization target through a subjective method and an entropy weight method, and unifying the three optimization targets into one comprehensive optimization target;
the electrical parameter optimal value acquisition module is used for acquiring a group of parameters with the minimum comprehensive optimization target through parameters in a trained deep belief network, wherein the group of parameters with the minimum comprehensive optimization target is the electrical parameter optimal value.
In a third aspect, the present invention provides a storage medium containing computer executable instructions which, when executed by a computer processor, implement the method of square wave voltage source electrical parameter optimization as described in the first aspect.
The invention provides a method, a system and a storage medium for square wave voltage source electrical parameter optimization, wherein the method for square wave voltage source electrical parameter optimization comprises the following steps: constructing a mathematical model of a square wave voltage source voltage forming loop, and solving through MATLAB to obtain a square wave voltage source state expression; determining electrical optimization parameters and optimization targets of a square wave voltage source voltage forming loop according to a square wave voltage source state expression, and determining a value range of the electrical optimization parameters by combining a mapping method and a load voltage expression; calculating to obtain a training sample set for training the network according to the value range of the electrical optimization parameters; training through a deep belief network according to a training sample set, and calculating the relation between each electrical parameter of a square wave voltage source voltage forming loop and an optimization target until the deep belief network meets the error requirement to obtain a trained deep belief network; unifying the optimization targets into a comprehensive optimization target; in the trained deep belief network, a group of parameters with the minimum comprehensive optimization target is obtained through the parameters, and the group of parameters with the minimum comprehensive optimization target is the optimal value of the electrical parameter. According to the square wave voltage source electrical parameter optimization method, the deep belief network is used, so that the test times and the test time are reduced, the deep belief network is trained firstly, namely the deep belief network is pre-trained, and the initial parameters of the network are more accurate.
The foregoing is merely a detailed description of the invention, and it should be noted that modifications and adaptations by those skilled in the art may be made without departing from the principles of the invention, and should be considered as within the scope of the invention.

Claims (6)

1. A method of electrical parameter optimization of a square wave voltage source, the method comprising:
s1: constructing a mathematical model of a square wave voltage source voltage forming loop, and solving through MATLAB to obtain a square wave voltage source state expression, wherein the square wave voltage source state expression comprises a direct current voltage source voltage expression, a load voltage expression, an energy storage capacitor expression and a load capacitor expression;
s2: determining electrical optimization parameters and optimization targets of a square wave voltage source voltage forming loop according to a square wave voltage source state expression, and determining the value range of the electrical optimization parameters by combining a mapping method and a load voltage expression, wherein the electrical optimization parameters comprise an energy storage capacitor, a current limiting resistor and a pull-down resistor, the optimization targets comprise the rise time, the duration and the maximum current of the square wave voltage, and the value range of the electrical optimization parameters comprises the value range of the energy storage capacitor, the value range of the current limiting resistor and the value range of the pull-down resistor;
s3: calculating to obtain a training sample set for training the network according to the value range of the electrical optimization parameters;
s4: training through a deep belief network according to a training sample set, and calculating the relation between each electrical parameter of a square wave voltage source voltage forming loop and an optimization target until the deep belief network meets the error requirement to obtain a trained deep belief network;
s5: unifying the three optimization targets into a comprehensive optimization target;
s6: in the trained deep belief network, a group of parameters with the minimum comprehensive optimization target is obtained through the parameters, and the group of parameters with the minimum comprehensive optimization target is the optimal value of the electrical parameter.
2. The method for optimizing electrical parameters of a square wave voltage source according to claim 1, wherein the step of calculating a training sample set for training a network according to a value range of the electrical optimization parameters comprises:
s31: selecting an input sample within the value range of the electrical optimization parameter;
s32: output samples under input samples, namely a training sample set for training the network, are obtained through MATLAB/SIMULINK simulation.
3. The method for electrical parameter optimization of a square wave voltage source of claim 2, wherein each of the input samples comprises an energy storage capacitance value, a current limiting resistance value, and a pull-down resistance value.
4. The method for electrical parameter optimization of a square wave voltage source according to claim 1, wherein the three optimization objectives are unified into one comprehensive optimization objective, and the comprehensive optimization objective is obtained by setting the weight of the optimization objective through an optimization objective subjective method and an entropy weight method.
5. The system for optimizing the electrical parameters of the square wave voltage source is characterized by comprising a square wave voltage source state expression obtaining module, an electrical optimization parameter-optimization target obtaining module, a training sample set obtaining module, a training module, a comprehensive optimization target obtaining module and an electrical parameter optimal value obtaining module, wherein:
the square wave voltage source state expression obtaining module is used for constructing a mathematical model of a square wave voltage source voltage forming loop, and solving through MATLAB to obtain a square wave voltage source state expression, wherein the square wave voltage source state expression comprises a direct current voltage source voltage expression, a load voltage expression, an energy storage capacitor expression and a load capacitor expression;
the electrical optimization parameter-optimization target acquisition module is used for determining electrical optimization parameters and optimization targets of a square wave voltage source voltage forming loop according to a square wave voltage source state expression, combining a mapping method and a load voltage expression to determine a value range of the electrical optimization parameters, wherein the electrical optimization parameters comprise an energy storage capacitor, a current limiting resistor and a pull-down resistor, the optimization targets comprise square wave voltage rise time, duration and maximum current, and the value range of the electrical optimization parameters comprises the value range of the energy storage capacitor, the value range of the current limiting resistor and the value range of the pull-down resistor;
the training sample set acquisition module is used for calculating a training sample set used for training the network according to the value range of the electrical optimization parameters;
the training module is used for training through a deep belief network according to a training sample set, calculating the relation between each electrical parameter of a square wave voltage source voltage forming loop and an optimization target until the deep belief network meets the error requirement, and obtaining the trained deep belief network;
the comprehensive optimization target obtaining module is used for setting the weight of the optimization target through a subjective method and an entropy weight method, and unifying the three optimization targets into one comprehensive optimization target;
the electrical parameter optimal value acquisition module is used for acquiring a group of parameters with the minimum comprehensive optimization target through parameters in a trained deep belief network, wherein the group of parameters with the minimum comprehensive optimization target is the electrical parameter optimal value.
6. A storage medium containing computer executable instructions which, when executed by a computer processor, implement a method of square wave voltage source electrical parameter optimization according to any of claims 1-4.
CN202010784036.9A 2020-08-06 2020-08-06 Method and system for optimizing electrical parameters of square wave voltage source and storage medium Active CN111881628B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202010784036.9A CN111881628B (en) 2020-08-06 2020-08-06 Method and system for optimizing electrical parameters of square wave voltage source and storage medium

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202010784036.9A CN111881628B (en) 2020-08-06 2020-08-06 Method and system for optimizing electrical parameters of square wave voltage source and storage medium

Publications (2)

Publication Number Publication Date
CN111881628A true CN111881628A (en) 2020-11-03
CN111881628B CN111881628B (en) 2023-01-20

Family

ID=73211862

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202010784036.9A Active CN111881628B (en) 2020-08-06 2020-08-06 Method and system for optimizing electrical parameters of square wave voltage source and storage medium

Country Status (1)

Country Link
CN (1) CN111881628B (en)

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110414718A (en) * 2019-07-04 2019-11-05 上海工程技术大学 A kind of distribution network reliability index optimization method under deep learning
CN110737998A (en) * 2019-09-25 2020-01-31 中国电力科学研究院有限公司 equalizing ring optimization design method based on finite element and depth belief network

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110414718A (en) * 2019-07-04 2019-11-05 上海工程技术大学 A kind of distribution network reliability index optimization method under deep learning
CN110737998A (en) * 2019-09-25 2020-01-31 中国电力科学研究院有限公司 equalizing ring optimization design method based on finite element and depth belief network

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
孔祥玉等: "基于深度信念网络的短期负荷预测方法", 《电力系统自动化》 *
李辉等: "基于改进深度信念网络的电力变压器故障诊断研究", 《软件导刊》 *

Also Published As

Publication number Publication date
CN111881628B (en) 2023-01-20

Similar Documents

Publication Publication Date Title
US11544917B2 (en) Power electronic circuit fault diagnosis method based on optimizing deep belief network
US20220222409A1 (en) Method and system for predicting remaining useful life of analog circuit
CN107085633B (en) Device and method for multi-point vibration response frequency domain prediction based on support vector machine
CN108646149A (en) Fault electric arc recognition methods based on current characteristic extraction
Houlian et al. State of charge prediction of supercapacitors via combination of Kalman filtering and backpropagation neural network
CN110889228B (en) Transformer oil paper insulation aging prediction method based on chicken swarm optimization BP neural network
CN110045317B (en) Mutual inductor metering error online detection method and system
CN112632850A (en) Method and system for detecting abnormal battery in lithium battery pack
CN103246762A (en) Method of comprehensive evaluation for simulation credibility of electric propulsion system
CN112014696A (en) Dry-type air-core reactor fault diagnosis method and device based on partial discharge characteristics
CN116990625B (en) Function switching system and method of intelligent quick-checking device of distribution transformer
CN117949884A (en) Power quality and voltage monitoring and calibrating method and system based on machine learning
Coban et al. Artificial neural network based fault location on 230 kv transmission line using voltage and current signals
CN117972460A (en) Operation fault discrimination method for high-voltage current transformer
CN111881628B (en) Method and system for optimizing electrical parameters of square wave voltage source and storage medium
CN117314236A (en) BP neural network-based power distribution equipment health state evaluation method and system
CN117110880A (en) Battery multi-state joint estimation method for cloud-edge end cooperation
CN108037424B (en) Power capacitor insulation state evaluation method and system based on multiple data sources
CN103942416B (en) Voltage estimation method based on weighted node spanning tree
CN110737998A (en) equalizing ring optimization design method based on finite element and depth belief network
CN110598849A (en) HMET scattering parameter extraction method and system based on neural network and storage medium
CN115291039A (en) Single-phase earth fault line selection method of resonance earthing system
CN114966224A (en) Improved neural network-based modular multilevel converter capacitance monitoring method
CN114019296A (en) Distribution line ground fault identification method based on BP neural network
CN113325317A (en) Power battery fault diagnosis method and system based on improved RBF neural network

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant