CN117411033B - Inertia damping control method and system based on radial neural network - Google Patents
Inertia damping control method and system based on radial neural network Download PDFInfo
- Publication number
- CN117411033B CN117411033B CN202311688232.6A CN202311688232A CN117411033B CN 117411033 B CN117411033 B CN 117411033B CN 202311688232 A CN202311688232 A CN 202311688232A CN 117411033 B CN117411033 B CN 117411033B
- Authority
- CN
- China
- Prior art keywords
- output
- error
- inertia
- damping
- representing
- 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.)
- Active
Links
- 238000013016 damping Methods 0.000 title claims abstract description 84
- 238000013528 artificial neural network Methods 0.000 title claims abstract description 53
- 238000000034 method Methods 0.000 title claims abstract description 14
- 238000012549 training Methods 0.000 claims abstract description 96
- 238000012360 testing method Methods 0.000 claims abstract description 25
- 238000011156 evaluation Methods 0.000 claims abstract description 15
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 10
- 230000001360 synchronised effect Effects 0.000 claims abstract description 10
- 238000012216 screening Methods 0.000 claims abstract description 9
- 238000004088 simulation Methods 0.000 claims abstract description 9
- 210000002569 neuron Anatomy 0.000 claims description 23
- 238000004364 calculation method Methods 0.000 claims description 4
- 238000012545 processing Methods 0.000 claims description 3
- 238000005070 sampling Methods 0.000 claims description 3
- 238000013461 design Methods 0.000 abstract description 5
- 230000001629 suppression Effects 0.000 abstract description 2
- 238000010586 diagram Methods 0.000 description 7
- 238000005516 engineering process Methods 0.000 description 5
- 230000000694 effects Effects 0.000 description 4
- 241000764238 Isis Species 0.000 description 2
- 230000003044 adaptive effect Effects 0.000 description 2
- 238000010248 power generation Methods 0.000 description 2
- 230000007547 defect Effects 0.000 description 1
- 230000005764 inhibitory process Effects 0.000 description 1
- 230000010355 oscillation Effects 0.000 description 1
- 230000035515 penetration Effects 0.000 description 1
- 230000035699 permeability Effects 0.000 description 1
- 230000009897 systematic effect Effects 0.000 description 1
Classifications
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J3/00—Circuit arrangements for ac mains or ac distribution networks
- H02J3/24—Arrangements for preventing or reducing oscillations of power in networks
- H02J3/241—The oscillation concerning frequency
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B13/00—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion
- G05B13/02—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric
- G05B13/0265—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric the criterion being a learning criterion
- G05B13/027—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric the criterion being a learning criterion using neural networks only
-
- 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
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02P—CONTROL OR REGULATION OF ELECTRIC MOTORS, ELECTRIC GENERATORS OR DYNAMO-ELECTRIC CONVERTERS; CONTROLLING TRANSFORMERS, REACTORS OR CHOKE COILS
- H02P9/00—Arrangements for controlling electric generators for the purpose of obtaining a desired output
- H02P9/10—Control effected upon generator excitation circuit to reduce harmful effects of overloads or transients, e.g. sudden application of load, sudden removal of load, sudden change of load
- H02P9/105—Control effected upon generator excitation circuit to reduce harmful effects of overloads or transients, e.g. sudden application of load, sudden removal of load, sudden change of load for increasing the stability
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J2203/00—Indexing scheme relating to details of circuit arrangements for AC mains or AC distribution networks
- H02J2203/20—Simulating, e g planning, reliability check, modelling or computer assisted design [CAD]
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02P—CONTROL OR REGULATION OF ELECTRIC MOTORS, ELECTRIC GENERATORS OR DYNAMO-ELECTRIC CONVERTERS; CONTROLLING TRANSFORMERS, REACTORS OR CHOKE COILS
- H02P2103/00—Controlling arrangements characterised by the type of generator
- H02P2103/20—Controlling arrangements characterised by the type of generator of the synchronous type
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Artificial Intelligence (AREA)
- Theoretical Computer Science (AREA)
- Evolutionary Computation (AREA)
- Health & Medical Sciences (AREA)
- General Physics & Mathematics (AREA)
- Software Systems (AREA)
- Computing Systems (AREA)
- Data Mining & Analysis (AREA)
- Power Engineering (AREA)
- Life Sciences & Earth Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Mathematical Physics (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- General Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Medical Informatics (AREA)
- Automation & Control Theory (AREA)
- Feedback Control In General (AREA)
- Control Of Electric Motors In General (AREA)
Abstract
The invention provides an inertia damping control method based on a radial neural network, which adopts improved self-adaptive algorithm simulation to obtain initialization parameters of a synchronous motor; the initialization parameters comprise angular speed, angular speed offset rate, output inertia and output damping; screening the initialization parameters and classifying the initialization parameters into training set parameters and testing set parameters; performing artificial radial neural network training on the training set parameters and the testing set parameters to obtain the output inertia and output damping values; and carrying out error evaluation on the numerical value of the output inertia and the output damping. The invention can improve the control precision and response speed, optimize parameter design and adjustment, and realize excellent dynamic performance such as smoother power output, smaller overshoot, smaller frequency deviation, better power fluctuation suppression and the like.
Description
Technical Field
The invention relates to the technical field of power grid control, in particular to an inertia damping control method based on a radial neural network.
Background
As the renewable energy power generation duty ratio is larger and larger, inertia and damping are smaller and smaller, and the stability and reliability of the operation of the power grid are seriously threatened. Renewable energy power generation has volatility and uncertainty, so that inertia and damping in a power grid are reduced, and the power grid frequency fluctuation is excessive, and even large-scale power outage is caused.
Grid frequency is an important indicator reflecting the supply and demand balance of the grid, and its stability depends on inertia and damping in the grid. The inertia may buffer power fluctuations in the power grid and the damping may dampen oscillations in the power grid.
Document "Self-adaptive virtual inertia control-based fuzzy logic to improve frequency stability of microgrid with high renewable penetration [ J ]. IEEE Access,2019,7:76071-83. (adaptive virtual inertia control based on fuzzy logic to improve high renewable permeability microgrid frequency stability [ J ]. IEEE Access journal, 2019, 7:76071-83.)" discloses an adaptive virtual inertia control system using fuzzy logic, which automatically adjusts virtual inertia constants to ensure frequency stability using input signals injected with real power and systematic frequency deviation. The scheme also optimizes the parameters of the virtual synchronous machine by utilizing an intelligent control technology, and realizes quick inertial response and stable frequency. However, the scheme has the defects that the design of the fuzzy control algorithm is completely based on experience, no clear theoretical basis exists, the rough problem of parameter design exists, and the influence of damping is not considered.
Disclosure of Invention
The invention provides an inertia damping control method based on a radial neural network, which comprises the following steps:
adopting improved self-adaptive algorithm simulation to obtain initialization parameters of the synchronous motor;
the initialization parameters comprise angular speed, angular speed offset rate, output inertia and output damping;
screening the initialization parameters and classifying the initialization parameters into training set parameters and testing set parameters;
performing artificial radial neural network training on the training set parameters and the testing set parameters to obtain the output inertia and output damping values;
and carrying out error evaluation on the numerical value of the output inertia and the output damping.
The artificial radial neural network is a radial basis function network (RBF), and has the following structure and connection mode:
(1) The input layer consists of two neurons, and receives power grid frequency deviation and power fluctuation as input signals respectively;
(2) The hidden layer consists of a plurality of neurons, each neuron corresponds to a Gaussian basis function, and the central value, standard deviation and weight of the basis function are adaptively adjusted according to the characteristic points of the input signals;
(3) The output layer is composed of two neurons, and virtual inertia and damping coefficients are respectively output as control signals.
It should be noted that, the step of performing artificial radial neural network training on the training set parameter and the test set parameter includes:
;
wherein,representing the i-th output inertia; />Representing the output damping of the ith, i represents a sequence number, and m represents the number of neurons of the current hidden layer;
wherein,is the i-th angular velocity, +.>Is the i-th angular velocity offset rate, < >>Is the standard deviation of the ith, +.>Is the base function center value of the ith.
The following is a description ofThe calculation formula of (2) is as follows: />;
Wherein the saidRepresenting the nearest cluster base +.>Indicate distance (I)>Representing the overlap factor.
The overlapping coefficient isIs a positive real number, is used for controlling the width of the Gaussian basis function, has the value range of (0, 1), and is determined by the following steps: />Wherein->Is the maximum distance between the cluster base points.
It should be noted that, the step of performing artificial radial neural network training on the training set parameter and the test set parameter further includes:
judging errors of k and k-1 training set parameters;
comparing the error of the training set parameter with the set parameter value;
the error of the training set parameter is smaller than the set parameter value, and then:
;
wherein,representing the i-th output inertia; />Representing the output damping of the ith, i represents a sequence number, and m represents the number of neurons of the current hidden layer;
wherein,is the i-th angular velocity, +.>Is the i-th angular velocity offset rate, < >>Is the standard deviation of the ith, +.>Is the base function center value of the ith.
It should be noted that, the step of performing artificial radial neural network training on the training set parameter and the test set parameter further includes:
judging errors of k and k-1 training set parameters;
it should be further noted that the meaning of k is the number of current training set parameters, and the meaning of k-1 is the number of last training set parameters. This step can therefore also be expressed as: and judging the error between the number k of the parameters of the current training set and the number k-1 of the parameters of the last training set.
Comparing the error of the training set parameter with the set parameter value;
the error of the training set parameter is larger than the set parameter value, and the value of the ith clustering base point and the value of the weight are adjusted;
further, the step of "adjusting the numerical value of the ith cluster base point and the weight" includes:
calculating an error functionWherein->For the number of neurons in the output layer, < > and->For the desired output value, +.>Is the actual output value;
calculating the weight and the i-th cluster base point pair error functionIs a partial derivative of (2);
updating the weight and the ith clustering base point according to the partial derivative;
repeating the above steps until the error functionReaching a minimum value or meeting a stop condition;
when the error of the training set parameter is smaller than the set parameter value, adjusting to the error of the training set parameter is smaller than the set parameter value, and performing:
;
wherein,representing the i-th output inertia; />Representing the output damping of the ith, i represents a sequence number, and m represents the number of neurons of the current hidden layer;
wherein,is the i-th angular velocity, +.>Is the i-th angular velocity offset rate, < >>Is the standard deviation of the ith, +.>Is the base function center value of the ith.
It should be noted that the step of performing error evaluation on the output inertia and the output damping value includes:
counting the error value of the output inertia;
counting the error value of the output damping;
judging whether the error rate of the output inertia is less than one thousandth;
and judging whether the error rate of the output damping is less than one thousandth.
An inertia damping control system based on a radial neural network, comprising:
the sampling module is used for adopting improved self-adaptive algorithm simulation to obtain initialization parameters of the synchronous motor;
the initialization parameters comprise angular speed, angular speed offset rate, output inertia and output damping;
the signal processing module is used for screening the initialization parameters and classifying the initialization parameters into training set parameters and testing set parameters;
the radial neural network control module is used for carrying out artificial radial neural network training on the training set parameters and the testing set parameters to obtain the output inertia and output damping values;
and the error evaluation module is used for evaluating the errors of the output inertia and the output damping value.
It should be noted that the radial neural network control module includes:
the judging unit is used for judging errors of k training set parameters and k-1 training set parameters;
the comparison unit is used for comparing the error of the training set parameter with the set parameter value;
the operation unit is used for carrying out the following steps if the error of the training set parameter is smaller than the set parameter value:
;
wherein,representing the i-th output inertia; />Representing the output damping of the ith, i represents a sequence number, and m represents the number of neurons of the current hidden layer;
wherein,is the i-th angular velocity, +.>Is the i-th angular velocity offset rate, < >>Is the standard deviation of the ith, +.>Is the base function center value of the ith.
It should be noted that the radial neural network control module includes:
the judging unit compares the error of the training set parameter with the set parameter value;
the comparison unit is used for adjusting the values of the ith clustering base point and the weight if the error of the training set parameter is larger than the set parameter value;
the adjusting unit is used for adjusting the training set parameters until the errors of the training set parameters are smaller than the set parameter values;
the arithmetic unit runs the following formula:
;
wherein,representing the i-th output inertia; />Representing the output damping of the ith, i represents a sequence number, and m represents the number of neurons of the current hidden layer;
wherein,is the i-th angular velocity, +.>Is the i-th angular velocity offset rate, < >>Is the standard deviation of the ith, +.>Is the base function center value of the ith.
It should be noted that the error evaluation module includes:
the statistics unit is used for counting the error value of the output inertia; the error value used for carrying out statistics on the output damping;
the error judging unit is used for judging whether the error rate of the output inertia is less than one thousandth; and judging whether the error rate of the output damping is less than one thousandth.
Compared with the prior art, the inertia damping control method based on the radial neural network, which is selected by the invention, adopts improved self-adaptive algorithm simulation to obtain the initialization parameters of the synchronous motor; the initialization parameters comprise angular speed, angular speed offset rate, output inertia and output damping; screening the initialization parameters and classifying the initialization parameters into training set parameters and testing set parameters; performing artificial radial neural network training on the training set parameters and the testing set parameters to obtain the output inertia and output damping values; and carrying out error evaluation on the numerical value of the output inertia and the output damping. The invention utilizes the radial neural network control technology to simultaneously consider the effects of inertia and damping, optimize the external characteristics and enhance the stability and reliability of the power grid, while the prior art does not consider the influence of damping; the control precision and response speed can be improved, the parameter design and adjustment are optimized, and the overshoot and the adjustment time in the prior art are larger, so that the control effect is poor. The invention can realize the excellent dynamic performance of smoother power output, smaller overshoot, smaller frequency deviation, better power fluctuation inhibition and the like by utilizing the radial neural network control technology; the method can also realize the characteristics of rapid convergence, high precision, low complexity, strong generalization capability and the like, and optimize network structures and parameters by adjusting clustering base points and weights.
Drawings
For a clearer description of the technical solutions of the embodiments of the present invention, the drawings that are needed in the description of the embodiments will be briefly introduced below, it being obvious that the drawings in the description below are only some embodiments of the present invention, and that other drawings can be obtained according to these drawings without inventive effort for a person skilled in the art, wherein:
FIG. 1 is a schematic flow chart of an inertia damping control method based on a radial neural network;
FIG. 2 is a flow chart illustrating an embodiment of the step S3 shown in FIG. 1;
FIG. 3 is a flowchart illustrating another embodiment of the step S3 shown in FIG. 1;
fig. 4 is a schematic flow chart of step S4 shown in fig. 1;
FIG. 5 is a schematic diagram of an inertia damping control system based on a radial neural network according to the present invention;
FIG. 6 is a schematic diagram of one embodiment of a radial neural network control module provided in FIG. 5;
FIG. 7 is a schematic diagram of another embodiment of the radial neural network control module provided in FIG. 5;
FIG. 8 is a schematic diagram of the error estimation module provided in FIG. 5.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and fully with reference to the accompanying drawings, in which it is evident that the embodiments described are only some, but not all embodiments of the invention. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
Referring to fig. 1, fig. 1 is a schematic flow chart of an inertia damping control method based on a radial neural network according to the present invention; comprising the following steps:
s1, adopting improved self-adaptive algorithm simulation to obtain initialization parameters of a synchronous motor;
s2, screening the initialization parameters and classifying the initialization parameters into training set parameters and testing set parameters;
s3, training the training set parameters and the testing set parameters by using an artificial radial neural network to obtain the output inertia and output damping values;
and S4, carrying out error evaluation on the output inertia and the output damping value.
It should be noted that, the step of performing artificial radial neural network training on the training set parameter and the test set parameter includes:
;
wherein,representing the i-th output inertia; />Representing the output damping of the ith, i represents a sequence number, and m represents the number of neurons of the current hidden layer;
wherein,is the i-th angular velocity, +.>Is the i-th angular velocity offset rate, < >>Is the standard deviation of the ith, +.>Is the base function center value of the ith.
The following is a description ofThe calculation formula of (2) is as follows: />;
Wherein the saidRepresenting the nearest cluster base +.>Indicate distance (I)>Representing the overlap factor.
The overlapping coefficient isIs a positive real number, is used for controlling the width of the Gaussian basis function, has the value range of (0, 1), and is determined by the following steps: />Wherein->Is the maximum distance between the cluster base points.
Referring to fig. 2, fig. 2 is a flow chart illustrating an embodiment of step S3 shown in fig. 1; comprising the following steps:
s31, judging errors of k training set parameters and k-1 training set parameters;
s32, comparing the error of the training set parameter with the set parameter value;
s33, the error of the training set parameter is smaller than the set parameter value;
s34, performing operation of the following formula:
；
wherein,representing the i-th output inertia; />Representing the output damping of the ith, i represents a sequence number, and m represents the number of neurons of the current hidden layer;
wherein,is the i-th angular velocity, +.>Is the i-th angular velocity offset rate, < >>Is the standard deviation of the ith, +.>Is the base function center value of the ith.
It should be further noted that the meaning of k is the number of current training set parameters, and the meaning of k-1 is the number of last training set parameters. This step can therefore also be expressed as: and judging the error between the number k of the parameters of the current training set and the number k-1 of the parameters of the last training set.
Referring to fig. 3, fig. 3 is a flow chart of another embodiment of step S3 shown in fig. 1; comprising the following steps:
s31, judging errors of k training set parameters and k-1 training set parameters;
s32, comparing the error of the training set parameter with the set parameter value;
s303, if the error of the training set parameter is larger than the set parameter value, the value of the ith clustering base point and the value of the weight are adjusted;
s304, adjusting the error of the training set parameters to be smaller than the set parameter values;
s34, performing operation of the following formula:
；
wherein,representing the i-th output inertia; />Representing the output damping of the ith, i represents a sequence number, and m represents the number of neurons of the current hidden layer;
wherein,is the i-th angular velocity, +.>Is the i-th angular velocity offset rate, < >>Is the standard deviation of the ith, +.>Is the base function center value of the ith.
Further, the step S33 includes:
s331, calculating an error function；
Wherein,for the number of neurons in the output layer, < > and->For the desired output value, +.>Is the actual output value;
s332, calculating the weight and the ith clustering base point pair error functionIs a partial derivative of (2);
s333, updating the weight and the ith clustering base point according to the partial derivative;
s334, repeating the above steps until the error functionReaching a minimum or meeting a stop condition.
As shown in fig. 4, fig. 4 is a schematic flow chart of step S4 shown in fig. 1;
comprising the following steps:
s41, counting the error value of the output inertia;
s42, counting the error value of the output damping;
s43, judging whether the error rate of the output inertia is less than one thousandth;
s44, judging whether the error rate of the output damping is smaller than one thousandth.
Referring to fig. 5, fig. 5 is a schematic diagram of an inertia damping control system based on a radial neural network according to the present invention; comprising the following steps:
the sampling module 1 is used for adopting improved self-adaptive algorithm simulation to obtain initialization parameters of the synchronous motor;
the initialization parameters comprise angular speed, angular speed offset rate, output inertia and output damping;
the signal processing module 2 is used for screening the initialization parameters and classifying the initialization parameters into training set parameters and testing set parameters;
the radial neural network control module 3 is used for performing artificial radial neural network training on the training set parameters and the testing set parameters to obtain the output inertia and output damping values;
and the error evaluation module 4 is used for performing error evaluation on the output inertia and the output damping value.
Referring to fig. 6, fig. 6 is a schematic diagram illustrating an embodiment of a radial neural network control module provided in fig. 5; the radial neural network control module includes:
a judging unit 31 for judging errors of k and k-1 training set parameters;
a comparing unit 32, configured to compare the error of the training set parameter with the set parameter value;
the operation unit 33 is configured to, when the error of the training set parameter is smaller than the set parameter value, perform:
；
wherein,representing the i-th output inertia; />Representing the output damping of the ith, i represents a sequence number, and m represents the number of neurons of the current hidden layer;
wherein,is the i-th angular velocity, +.>Is the i-th angular velocity offset rate, < >>Is the standard deviation of the ith, +.>Is the base function center value of the ith.
Referring to fig. 7, fig. 7 is a schematic diagram of another embodiment of the radial neural network control module provided in fig. 5;
the radial neural network control module includes:
a judging unit 31 for comparing the error of the training set parameter with the set parameter value;
a comparison unit 32, configured to adjust the values of the i-th clustering base point and the weight if the error of the training set parameter is greater than the set parameter value;
an adjusting unit 302, configured to adjust the training set parameter to an error smaller than the set parameter value;
the arithmetic unit 33 runs the following formula:
；
wherein,representing the i-th output inertia; />Representing the output damping of the ith, i represents a sequence number, and m represents the number of neurons of the current hidden layer;
wherein,is the i-th angular velocity, +.>Is the i-th angular velocity offset rate, < >>Is the standard deviation of the ith, +.>Is the base function center value of the ith.
It should be noted that the error evaluation module includes:
a statistics unit 41, configured to perform statistics on an error value of the output inertia; the error value used for carrying out statistics on the output damping;
an error judging unit 42, configured to judge whether the error rate of the output inertia is less than one thousandth; and judging whether the error rate of the output damping is less than one thousandth.
Compared with the prior art, the inertia damping control method based on the radial neural network, which is selected by the invention, adopts improved self-adaptive algorithm simulation to obtain the initialization parameters of the synchronous motor; the initialization parameters comprise angular speed, angular speed offset rate, output inertia and output damping; screening the initialization parameters and classifying the initialization parameters into training set parameters and testing set parameters; performing artificial radial neural network training on the training set parameters and the testing set parameters to obtain the output inertia and output damping values; and carrying out error evaluation on the numerical value of the output inertia and the output damping. The invention can simultaneously consider the effects of inertia and damping by utilizing the radial neural network control technology, optimize the external characteristics and enhance the stability and reliability of the power grid, while the prior art does not consider the influence of damping. The control precision and response speed can be improved, the parameter design and adjustment are optimized, and the overshoot and the adjustment time in the prior art are larger, so that the control effect is poor. The invention can realize the excellent dynamic performance of smoother power output, smaller overshoot, smaller frequency deviation, better power fluctuation suppression and the like by utilizing the radial neural network control technology.
Claims (6)
1. The inertia damping control method based on the radial neural network is characterized by comprising the following steps of:
adopting improved self-adaptive algorithm simulation to obtain initialization parameters of the synchronous motor;
the initialization parameters comprise angular speed, angular speed offset rate, output inertia and output damping;
screening the initialization parameters and classifying the initialization parameters into training set parameters and testing set parameters;
performing artificial radial neural network training on the training set parameters and the testing set parameters to obtain the output inertia and output damping values;
performing error evaluation on the output inertia and the output damping value;
the step of performing artificial radial neural network training on the training set parameters and the test set parameters comprises the following steps:
；
wherein,representing the firstiThe output inertia of each; />Representing the firstiThe output damping of each of the plurality of the output dampers,iindicating a sequence number,mrepresenting the number of neurons of the current hidden layer;
wherein,is the firstiAngular velocity of>Is the firstiAngular velocity offset of (a),>is the firstiStandard deviation of->Is the firstiA base function center value of each;
the saidThe calculation formula of (2) is as follows: />；
Wherein the saidRepresenting the nearest cluster base +.>Indicate distance (I)>Representing the overlap factor;
the step of error evaluating the output inertia and the output damping value includes:
counting the error value of the output inertia;
counting the error value of the output damping;
judging whether the error rate of the output inertia is less than one thousandth;
and judging whether the error rate of the output damping is less than one thousandth.
2. The radial neural network-based inertia damping control method of claim 1, wherein the step of artificial radial neural network training of the training set parameters and the test set parameters further comprises:
judgingkSum of allk-Error of 1 training set parameter;
comparing the error of the training set parameter with the set parameter value;
the error of the training set parameter is smaller than the set parameter value, and then:
；
wherein,representing the firstiThe output inertia of each; />Representing the firstiThe output damping of each of the plurality of the output dampers,iindicating a sequence number,mrepresenting the number of neurons of the current hidden layer;
wherein,is the firstiAngular velocity of>Is the firstiAngular velocity offset of (a),>is the firstiStandard deviation of->Is the firstiAnd the base function center value of each.
3. The radial neural network-based inertia damping control method of claim 1, wherein the step of artificial radial neural network training of the training set parameters and the test set parameters further comprises:
judgment of the firstkAnd (b)k-Error of 1 training set parameter;
comparing the error of the training set parameter with the set parameter value;
the error of the training set parameter is larger than the set parameter value, and the first is adjustedi numberThe numerical values of the clustering base points and the weights;
the error of the training set parameter is adjusted to be smaller than the set parameter value;
further, the step of adjusting to the training set parameter error smaller than the set parameter value includes:
calculating an error function；
Wherein,for the number of neurons in the output layer, < > and->For the desired output value, +.>Is the actual output value;
calculate the weight and the thi numberClustering base point pair error functionIs a partial derivative of (2);
updating the weight sum based on the partial derivativei numberClustering base points;
repeating the above steps until the error functionReaching a minimum value or meeting a stop condition;
when the error of the training set parameter is smaller than the set parameter value, the method is carried out:
；
wherein,representing the firstiThe output inertia of each; />Representing the firstiThe output damping of each of the plurality of the output dampers,iindicating a sequence number,mrepresenting the number of neurons of the current hidden layer;
wherein,is the firstiAngular velocity of>Is the firstiAngular velocity offset of (a),>is the firstiStandard deviation of->Is the firstiAnd the base function center value of each.
4. An inertia damping control system based on a radial neural network, comprising:
the sampling module is used for adopting improved self-adaptive algorithm simulation to obtain initialization parameters of the synchronous motor;
the initialization parameters comprise angular speed, angular speed offset rate, output inertia and output damping;
the signal processing module is used for screening the initialization parameters and classifying the initialization parameters into training set parameters and testing set parameters;
the radial neural network control module is used for carrying out artificial radial neural network training on the training set parameters and the testing set parameters to obtain the output inertia and output damping values;
the radial neural network control module operates as follows:
；
wherein,representing the firstiThe output inertia of each; />Representing the firstiThe output damping of each of the plurality of the output dampers,iindicating a sequence number,mrepresenting the number of neurons of the current hidden layer;
wherein,is the firstiAngular velocity of>Is the firstiAngular velocity offset of (a),>is the firstiStandard deviation of->Is the firstiA base function center value of each;
the saidThe calculation formula of (2) is as follows: />；
Wherein the saidRepresenting the nearest cluster base +.>Indicate distance (I)>Representing the overlap factor;
the error evaluation module is used for performing error evaluation on the output inertia and the output damping value;
the error evaluation module includes:
the statistics unit is used for counting the error value of the output inertia; the error value used for carrying out statistics on the output damping;
the error judging unit is used for judging whether the error rate of the output inertia is less than one thousandth; and judging whether the error rate of the output damping is less than one thousandth.
5. The radial neural network based inertia damping control system of claim 4, wherein the radial neural network control module comprises:
a judging unit for judgingkSum of allk-Error of 1 training set parameter;
the comparison unit is used for comparing the error of the training set parameter with the set parameter value;
the operation unit is used for carrying out the following steps if the error of the training set parameter is smaller than the set parameter value:
；
wherein,representing the firstiThe output inertia of each; />Representing the firstiThe output damping of each of the plurality of the output dampers,iindicating a sequence number,mrepresenting the number of neurons of the current hidden layer;
wherein,is the firstiAngular velocity of>Is the firstiAngular velocity offset of (a),>is the firstiStandard deviation of->Is the firstiAnd the base function center value of each.
6. The radial neural network based inertia damping control system of claim 4, wherein the radial neural network control module comprises:
the judging unit compares the error of the training set parameter with the set parameter value;
a comparison unit for adjusting the first parameter if the error of the training set parameter is larger than the set parameter valuei numberThe numerical values of the clustering base points and the weights;
the adjusting unit is used for adjusting the training set parameters until the errors of the training set parameters are smaller than the set parameter values;
the arithmetic unit runs the following formula:
；
wherein,representing the firstiThe output inertia of each; />Representing the firstiThe output damping of each of the plurality of the output dampers,iindicating a sequence number,mrepresenting the number of neurons of the current hidden layer;
wherein,is the firstiAngular velocity of>Is the firstiAngular velocity offset of (a),>is the firstiStandard deviation of->Is the firstiAnd the base function center value of each.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311688232.6A CN117411033B (en) | 2023-12-11 | 2023-12-11 | Inertia damping control method and system based on radial neural network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311688232.6A CN117411033B (en) | 2023-12-11 | 2023-12-11 | Inertia damping control method and system based on radial neural network |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117411033A CN117411033A (en) | 2024-01-16 |
| CN117411033B true CN117411033B (en) | 2024-03-01 |
Family
ID=89491108
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311688232.6A Active CN117411033B (en) | 2023-12-11 | 2023-12-11 | Inertia damping control method and system based on radial neural network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117411033B (en) |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113469332A (en) * | 2021-06-28 | 2021-10-01 | 上海电机学院 | Virtual synchronous generator inertia damping self-adaptive control method based on fuzzy nerves |
| CN115102188A (en) * | 2022-06-27 | 2022-09-23 | 湖南工学院 | VSG inertia and damping self-adaptive control method, system and computer readable medium |
| CN115912393A (en) * | 2022-10-18 | 2023-04-04 | 上海电力大学 | Multi-machine parallel VSG system stability improving method based on RBF neural network |
-
2023
- 2023-12-11 CN CN202311688232.6A patent/CN117411033B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113469332A (en) * | 2021-06-28 | 2021-10-01 | 上海电机学院 | Virtual synchronous generator inertia damping self-adaptive control method based on fuzzy nerves |
| CN115102188A (en) * | 2022-06-27 | 2022-09-23 | 湖南工学院 | VSG inertia and damping self-adaptive control method, system and computer readable medium |
| CN115912393A (en) * | 2022-10-18 | 2023-04-04 | 上海电力大学 | Multi-machine parallel VSG system stability improving method based on RBF neural network |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117411033A (en) | 2024-01-16 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104595106B (en) | Wind-power generating variable pitch control method based on intensified learning compensation | |
| CN104734175B (en) | A kind of intelligent correction method for realizing Wind turbines wind speed power curve | |
| CN112183641B (en) | Transient frequency stability assessment method and system integrating prediction-correction deep learning | |
| CN112152221B (en) | Load frequency control device and method suitable for information uncertainty system | |
| CN110374804A (en) | A kind of variable pitch control method based on the compensation of depth deterministic policy gradient | |
| CN112821420B (en) | XGboost-based prediction method and system for dynamic damping factor and multidimensional frequency index in ASFR model | |
| CN112821424B (en) | Power system frequency response analysis method based on data-model fusion driving | |
| CN109742773B (en) | Self-adaptive wide area damping controller | |
| CN112149883A (en) | Photovoltaic power prediction method based on FWA-BP neural network | |
| CN108551175A (en) | Power distribution network energy accumulation capacity configuration | |
| CN112615364A (en) | Novel wide-area intelligent cooperative control method for power grid stability control device | |
| CN113239547B (en) | Core backbone network frame construction method based on TLBO algorithm | |
| CN109376939B (en) | Power grid stability real-time prediction method based on adaptive neural network | |
| CN117411033B (en) | Inertia damping control method and system based on radial neural network | |
| CN116845886B (en) | Multi-port autonomous photovoltaic system network construction control method based on model prediction | |
| CN111969624A (en) | Damping control method and system of wind power grid-connected system containing virtual synchronous generator | |
| CN117526355A (en) | Distributed photovoltaic and adjustable load participation grid frequency modulation method and system | |
| CN116896093A (en) | Online analysis and optimization method for grid-connected oscillation stability of wind farm | |
| CN111224427A (en) | Micro-grid self-adaptive anti-islanding disturbance load impedance value prediction method and device | |
| CN112366733B (en) | Method, device, system and medium for optimizing PID (proportion integration differentiation) parameters of speed regulator of hydroelectric generating set | |
| CN112290539A (en) | Method and system for predicting transient voltage stability margin of power system | |
| Andreoiu et al. | Lyapunov's method based genetic algorithm for multi-machine PSS tuning | |
| Kamarzarrin et al. | A new fault-tolerant control of wind turbine pitch system based on ANN model and robust and optimal development of MRAC method | |
| CN110556873A (en) | VSG self-adaptive rotary inertia control method based on penalty function | |
| CN115617033B (en) | Ship formation method, system, device and storage medium |
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 |