CN108428021B - Micro-grid short-term load prediction model based on HSA-RRNN - Google Patents

Micro-grid short-term load prediction model based on HSA-RRNN Download PDF

Info

Publication number
CN108428021B
CN108428021B CN201810487990.4A CN201810487990A CN108428021B CN 108428021 B CN108428021 B CN 108428021B CN 201810487990 A CN201810487990 A CN 201810487990A CN 108428021 B CN108428021 B CN 108428021B
Authority
CN
China
Prior art keywords
layer
prediction model
neural network
hidden layer
nodes
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
Application number
CN201810487990.4A
Other languages
Chinese (zh)
Other versions
CN108428021A (en
Inventor
撖奥洋
于立涛
张滨
周生奇
王萍
钟世民
王黎
管春伟
张智晟
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.)
QINGDAO POWER SUPPLY Co OF STATE GRID SHANDONG ELECTRIC POWER Co
Qingdao University
Original Assignee
QINGDAO POWER SUPPLY Co OF STATE GRID SHANDONG ELECTRIC POWER Co
Qingdao University
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 QINGDAO POWER SUPPLY Co OF STATE GRID SHANDONG ELECTRIC POWER Co, Qingdao University filed Critical QINGDAO POWER SUPPLY Co OF STATE GRID SHANDONG ELECTRIC POWER Co
Priority to CN201810487990.4A priority Critical patent/CN108428021B/en
Publication of CN108428021A publication Critical patent/CN108428021A/en
Application granted granted Critical
Publication of CN108428021B publication Critical patent/CN108428021B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
    • G06Q50/06—Energy or water supply

Landscapes

  • Business, Economics & Management (AREA)
  • Engineering & Computer Science (AREA)
  • Economics (AREA)
  • Human Resources & Organizations (AREA)
  • Strategic Management (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Health & Medical Sciences (AREA)
  • Marketing (AREA)
  • General Physics & Mathematics (AREA)
  • General Business, Economics & Management (AREA)
  • Tourism & Hospitality (AREA)
  • Quality & Reliability (AREA)
  • Game Theory and Decision Science (AREA)
  • Operations Research (AREA)
  • Development Economics (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Public Health (AREA)
  • Water Supply & Treatment (AREA)
  • General Health & Medical Sciences (AREA)
  • Primary Health Care (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

The invention discloses a micro-grid short-term load prediction model based on HSA-RRNN, which is composed of a ridge wave recurrent neural network prediction model and comprises an input layer, a hidden layer, an associated layer and an output layer, wherein each neuron node in the hidden layer is correspondingly connected with the nodes of the associated layer one by one, and the weight between the nodes of the associated layer and the nodes of the hidden layer can be dynamically adjusted; and parameters in the ridge wave recurrent neural network prediction model are optimized by using the improved harmony search optimization algorithm, so that the method has the advantages of high convergence speed, stronger optimization capability and the like, and has better generalization and convergence. Through the prediction simulation and test of the actual micro-grid load, the prediction accuracy of the provided prediction model can be effectively improved.

Description

Micro-grid short-term load prediction model based on HSA-RRNN
Technical Field
The invention relates to a micro-grid short-term load prediction model based on HSA-RRNN.
Background
The smart city is a deep expansion and integrated application of information technology, is an important component of development of new industry of global strategy, provides a platform for development of a micro-grid group under advanced technical conditions, and provides powerful energy support for the smart city through development of the micro-grid group. The research on the source-network-load-storage cooperative optimization scheduling technology of the smart city micro-grid group has important practical significance. The microgrid load prediction is the basis of research on a microgrid group 'source-grid-load-storage' collaborative optimization scheduling technology, and accurate microgrid load prediction provides necessary basis for collaborative optimization scheduling. The micro-grid consists of a micro source, a load, an energy storage device, power electronic control protection equipment and the like, and the safety and the economy of the system are directly influenced by the prediction precision.
The neural network (RNN) is generated by simulating the visual cortex of the human brain, and its hidden layer neurons can receive data information in a specific direction and make the network output expected data through analysis processing. Compared with other traditional neural network models, the RNN model adopts the ridgelet function as the excitation function of the neurons in the hidden layer of the network model, has directional selectivity, can enable the network to contain more dimension information, thereby well processing data information with higher dimension and having better effect on approximation of nonlinear high-dimension functions.
In 1998, the ridge function was first proposed in doctor's paper by Candes, and its definition can be described in detail as: let the smoothing function ψ: rdFourier transform corresponding to → R
Figure GDA0003188300770000011
The following tolerance conditions were met:
Figure GDA0003188300770000012
let ψ be the admissible function, where ω is the argument and d is the spatial dimension. The ridge function ψ γ generated by the allowable function ψ satisfying the above condition is called a ridge wave function, and its expression is:
Figure GDA0003188300770000013
in formula (2), γ is a parameter space, i.e.:
γ＝(a,u,b)；a,b∈R；a＞0；u∈Sd-1 (9)
wherein, the scale vector of the ridge wave function is represented by a, the direction vector of the ridge wave function is represented by u, the position vector of the ridge wave function is represented by b, Sd-1 is represented as d-1 dimensional space, and | | | u | | | 1.
The RNN is named according to the excitation function of the neuron, namely, the ridgelet function is taken as the excitation function of the neuron of the hidden layer of the network model, the structure of the RNN is similar to that of the traditional feedforward neural network, and the RNN is composed of three layers of structures, namely an input layer, a hidden layer and an output layer.
The parameters in the existing ridge wave neural network prediction model need to be optimized, the parameters are selected without any theoretical basis, and the selection of the parameters has great influence on the prediction performance.
Disclosure of Invention
In order to solve the technical problems, the invention provides a micro-grid short-term load prediction model based on HSA-RRNN, so as to achieve the purposes of improving the prediction precision and meeting the actual scheduling prediction requirement.
In order to achieve the purpose, the technical scheme of the invention is as follows:
a microgrid short-term load prediction model based on HSA-RRNN is composed of a ridge wave recurrent neural network prediction model, wherein a ridge wave recurrent neural network is formed by adding an associated layer on the basis of the RNN prediction model, so that each neuron node in a hidden layer is correspondingly connected with the nodes of the associated layer one by one, and the weights between the nodes of the associated layer and the nodes of the hidden layer can be dynamically adjusted; and parameters in the spine wave recurrent neural network prediction model are optimized by using an improved harmony search optimization algorithm (HSA), wherein the optimization formula is as follows:
Figure GDA0003188300770000021
in the formula, PAR is disturbance probability, BW is disturbance bandwidth, N is current iteration number, N is maximum iteration number, r belongs to (0,1), and Z is any number.
In the above scheme, the mathematical expression of the prediction model of the ridge wave recurrent neural network is as follows:
Figure GDA0003188300770000022
Figure GDA0003188300770000023
Figure GDA0003188300770000024
Figure GDA0003188300770000031
in the formula, xi(i ═ 1,2, …, m) represents the input to the network;
Figure GDA0003188300770000032
representing the internal state of the hidden layer neuron node h, with inputs derived from the output states of the input layer and associated layer nodes;
Figure GDA0003188300770000033
representing nerves in the hidden layerThe output state of the metanode h;
Figure GDA0003188300770000034
an output value representing a model network;
Figure GDA0003188300770000035
represents the internal state of node c in the association layer;
Figure GDA0003188300770000036
representing the output state of the node c in the association layer; the admissible function used for the ridge function is:
Figure GDA0003188300770000037
where z represents the argument of the admissible function.
In the above scheme, the associated layer node may store the current output state of the hidden layer neuron node corresponding to the associated layer node, and transmit the current output state to each hidden layer neuron at the next time, which belongs to state feedback inside the model, and then perform repeated iterative update, thereby forming a dynamic memory function specific to the recurrent neural network.
In the above scheme, the input vector of the prediction model of the ridge wave recurrent neural network is [ x ]1，x2，…，xm]And m dimensions, namely a vector formed by related characteristic elements captured from the load sequence, the meteorological factors and the day type information in a time sequence order.
In the above scheme, the number of the hidden layer neuron nodes in the ridge wave recurrent neural network prediction model is p, and each node acquires input quantity from the input layer and the associated layer and uses the ridge wave function psiγCarrying out nonlinear transformation, and transmitting the output value to an output layer and a related layer; wherein the number of nodes of the associated layer is the same as the number of neuron nodes in the hidden layer.
In the above scheme, the number of the output layer nodes is 1, and in the structure of the prediction model of the ridge wave recurrent neural network, w is1o，……，whoFor adjustable weights of the connection between the hidden layer and the output layer, wc1，……，wchIs an adjustable weight value of the connection between the hidden layer and the associated layer.
Through the technical scheme, the HSA-RRNN-based microgrid short-term load prediction model is composed of a ridge wave recurrent neural network, parameters are optimized by using an improved harmony search optimization algorithm, and the method has the advantages of high convergence speed, higher optimization capability and the like, and has better generalization and convergence. Through the prediction simulation and test of the actual micro-grid load, the prediction accuracy of the provided prediction model can be effectively improved.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below.
Fig. 1 is a schematic structural diagram of a prediction model of a ridge wave recurrent neural network disclosed in the embodiment of the present invention.
Detailed Description
The technical solution in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention.
The invention provides a micro-grid short-term load prediction model based on HSA-RRNN, which comprises the following specific embodiments:
as shown in fig. 1, in the prediction model of the ridge wave recurrent neural network, the associated layer node can store the current output state of the corresponding hidden layer neuron node, and transmit the current output state to each hidden layer neuron at the next time, which belongs to the state feedback in the model, and then repeatedly update the state feedback, thereby forming the specific dynamic memory function of the recurrent neural network.
The input vector of the prediction model of the ridge wave recurrent neural network is [ x ]1，x2，…，xm]And m dimensions, namely a vector formed by related characteristic elements captured from the load sequence, the meteorological factors and the day type information in a time sequence order.
The number of hidden layer neuron nodes in the ridge wave recurrent neural network prediction model is p, each node acquires input quantity from an input layer and an associated layer, and the input quantity is obtained through a ridge wave function psiγCarrying out nonlinear transformation, and transmitting the output value to an output layer and a related layer; wherein the number of nodes of the associated layer is the same as the number of neuron nodes in the hidden layer.
The number of output layer nodes is 1, and in the prediction model structure of the ridge wave recurrent neural network, w1o，……，whoFor adjustable weights of the connection between the hidden layer and the output layer, wc1，……，wchIs an adjustable weight value of the connection between the hidden layer and the associated layer.
The mathematical expression of the prediction model of the ridge wave recurrent neural network is as follows:
Figure GDA0003188300770000041
Figure GDA0003188300770000042
Figure GDA0003188300770000043
Figure GDA0003188300770000044
in the formula, xi(i ═ 1,2, …, m) represents the input to the network;
Figure GDA0003188300770000045
representing the internal state of the hidden layer neuron node h, with inputs derived from the output states of the input layer and associated layer nodes;
Figure GDA0003188300770000046
representing the output state of a neuron node h in the hidden layer;
Figure GDA0003188300770000047
an output value representing a model network;
Figure GDA0003188300770000048
represents the internal state of node c in the association layer;
Figure GDA0003188300770000049
representing the output state of the node c in the association layer; the admissible function used for the ridge function is:
Figure GDA0003188300770000051
where z represents the argument of the admissible function.
HSA is a new heuristic global search algorithm, and simulates the process that a band plays the most beautiful harmony sound by repeatedly adjusting musical instruments, wherein each musical instrument is similar to each solving variable of an optimization problem, and the played harmony sound is similar to an objective function of the optimization problem. The algorithm first generates a Harmony Memory (HM) containing M solution vectors and feasible domains of variables, then randomly searches new solutions for the variables in the HM with a harmony library retention probability P1, finds new solutions in the feasible domains with probabilities 1-P1, locally perturbs the new solutions from the HM with a probability P2, finally compares the new solutions with the worst solution in the HM, and if the worst solution is better, replaces, and loops repeatedly until the maximum number of iterations is satisfied. Compared with the traditional algorithms such as particle swarm and heredity, the HSA has strong universality, high convergence rate and stronger optimizing capability.
The generation of new solutions in HSA and their perturbation are the most critical steps in which the present invention improves the relevant parameters. The basic HSA perturbs the new solution by a fixed probability PAR, the perturbation bandwidth is a fixed value BW, and the perturbation formula is xnewThe disturbance mode with fixed parameters can not better reflect the global search and local search capability of HSA, therefore, the invention adopts a disturbance probability and bandwidth along with the iteration timesThe mode of adaptation reduction, the concrete formula is:
Figure GDA0003188300770000052
in the formula, N is the current iteration number, N is the maximum iteration number, r belongs to (0,1), and Z is an arbitrary number.
After improvement, the PAR and BW of HSA are valued greatly in the early stage of the algorithm to enlarge the exploration range of the algorithm and obtain stronger global search capability, and the PAR and BW are gradually reduced along with the increase of iteration times, so that the algorithm is explored finely near an optimal value to obtain stronger local search capability.
In order to verify the effectiveness of the HSA-RRNN-based microgrid short-term load prediction model (model I), a BP-NN prediction model (model II) and a conventional RNN prediction model (model III) are adopted to predict 24-point loads of a microgrid at a certain day, and the prediction errors of the three models are compared and shown in Table 1.
TABLE 1 prediction error comparison of three prediction models
Figure GDA0003188300770000053
As can be seen from Table 1, the average absolute error is 9.22% and the maximum relative error is 19.71% using the conventional BP-NN prediction model. By adopting the conventional RNN prediction model, the average absolute error is reduced by 2.27 percent compared with the conventional BP-NN prediction model, and the maximum relative error is reduced by 3.51 percent. By adopting the HSA-RRNN-based micro-grid short-term load prediction model provided by the invention, the average absolute error is reduced by 4.13% compared with that of the traditional BP-NN prediction model, and the maximum relative error is reduced by 6.55%. The method is the most ideal of all models, and the result shows that the provided HSA-RRNN-based microgrid short-term load prediction model can effectively improve the prediction accuracy and meet the actual scheduling prediction requirement.
The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the invention. Thus, the present invention is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims (5)

1. The micro-grid short-term load prediction model based on HSA-RRNN is characterized by being composed of a ridge wave recurrent neural network prediction model and comprising an input layer, a hidden layer, an associated layer and an output layer, wherein each neuron node in the hidden layer is correspondingly connected with the associated layer node one by one, and the weight between the associated layer node and the hidden layer node can be dynamically adjusted; and parameters in the ridge wave recurrent neural network prediction model are optimized by using an improved harmony search optimization algorithm, wherein the optimization formulas of PAR and BW in the improved harmony search optimization algorithm are as follows:
Figure FDA0003188300760000011
in the formula, PAR is disturbance probability, BW is disturbance bandwidth, N is current iteration times, N is maximum iteration times, r is an element (0,1), and Z is an arbitrary number;
the mathematical expression of the prediction model of the ridge wave recurrent neural network is as follows:
Figure FDA0003188300760000012
Figure FDA0003188300760000013
Figure FDA0003188300760000014
Figure FDA0003188300760000015
in the formula, xiRepresents the input to the network, i ═ 1,2, …, m;
Figure FDA0003188300760000016
representing the internal state of the hidden layer neuron node h, with inputs derived from the output states of the input layer and associated layer nodes;
Figure FDA0003188300760000017
representing the output state of a neuron node h in the hidden layer;
Figure FDA0003188300760000018
an output value representing a model network;
Figure FDA0003188300760000019
represents the internal state of node c in the association layer;
Figure FDA00031883007600000110
representing the output state of the node c in the association layer; the admissible function used for the ridge function is:
Figure FDA00031883007600000111
where z represents the argument of the admissible function.
2. The HSA-RRNN-based microgrid short-term load prediction model of claim 1, wherein the associated layer nodes can store the current output states of the hidden layer neuron nodes corresponding thereto, transmit the current output states to each hidden layer neuron at the next time, and perform state feedback inside the model, and perform repeated iterative update, thereby forming a dynamic memory function specific to the recurrent neural network.
3. The HSA-RRNN-based microgrid short-term load prediction model of claim 1, wherein an input vector of the ridge wave recurrent neural network prediction model is [ x [ ]1，x2，…，xm]And m dimensions, namely a vector formed by related characteristic elements captured from the load sequence, the meteorological factors and the day type information in a time sequence order.
4. The HSA-RRNN-based microgrid short-term load prediction model of claim 1, wherein the hidden layer neuron nodes in the ridgelet recurrent neural network prediction model are p, and each node takes input from an input layer and an associated layer and passes through a ridgelet function ψγCarrying out nonlinear transformation, and transmitting the output value to an output layer and a related layer; wherein the number of nodes of the associated layer is the same as the number of neuron nodes in the hidden layer.
5. The HSA-RRNN-based microgrid short-term load prediction model of claim 1, wherein the number of output layer nodes is 1, w is in a ridge wave recurrent neural network prediction model structure1o，……，whoFor adjustable weights of the connection between the hidden layer and the output layer, wc1，……，wchIs an adjustable weight value of the connection between the hidden layer and the associated layer.
CN201810487990.4A 2018-05-21 2018-05-21 Micro-grid short-term load prediction model based on HSA-RRNN Active CN108428021B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201810487990.4A CN108428021B (en) 2018-05-21 2018-05-21 Micro-grid short-term load prediction model based on HSA-RRNN

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201810487990.4A CN108428021B (en) 2018-05-21 2018-05-21 Micro-grid short-term load prediction model based on HSA-RRNN

Publications (2)

Publication Number Publication Date
CN108428021A CN108428021A (en) 2018-08-21
CN108428021B true CN108428021B (en) 2021-10-12

Family

ID=63163550

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201810487990.4A Active CN108428021B (en) 2018-05-21 2018-05-21 Micro-grid short-term load prediction model based on HSA-RRNN

Country Status (1)

Country Link
CN (1) CN108428021B (en)

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110348631A (en) * 2019-07-09 2019-10-18 武汉四创自动控制技术有限责任公司 A kind of regionality Methods of electric load forecasting and system
CN112287982A (en) * 2020-10-14 2021-01-29 深圳大学 Data prediction method and device and terminal equipment

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102521671A (en) * 2011-11-29 2012-06-27 华北电力大学 Ultrashort-term wind power prediction method
CN104700116A (en) * 2015-03-13 2015-06-10 西安电子科技大学 Polarized SAR (synthetic aperture radar) image object classifying method based on multi-quantum ridgelet representation
CN105069468A (en) * 2015-07-28 2015-11-18 西安电子科技大学 Hyper-spectral image classification method based on ridgelet and depth convolution network
CN106529570A (en) * 2016-10-14 2017-03-22 西安电子科技大学 Image classification method based on deep ridgelet neural network

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102521671A (en) * 2011-11-29 2012-06-27 华北电力大学 Ultrashort-term wind power prediction method
CN104700116A (en) * 2015-03-13 2015-06-10 西安电子科技大学 Polarized SAR (synthetic aperture radar) image object classifying method based on multi-quantum ridgelet representation
CN105069468A (en) * 2015-07-28 2015-11-18 西安电子科技大学 Hyper-spectral image classification method based on ridgelet and depth convolution network
CN106529570A (en) * 2016-10-14 2017-03-22 西安电子科技大学 Image classification method based on deep ridgelet neural network

Non-Patent Citations (4)

* Cited by examiner, † Cited by third party
Title
A new wind power prediction method based on ridgelet transforms, hybrid feature selection and closed-loop forecasting;Hua等;《Advanced Engineering Informatics》;20180430;全文 *
Short-term wind power forecasting using ridgelet neural network;Nima等;《Electric Power Systems Research》;20111130;全文 *
基于分段分层相似日搜索和自适应脊波神经网络的风电功率多步预测;张宜阳等;《电网与清洁能源》;20150430;全文 *
基于脊波神经网络的短期风电功率预测;茆美琴等;《电力***自动化》;20110410;全文 *

Also Published As

Publication number Publication date
CN108428021A (en) 2018-08-21

Similar Documents

Publication Publication Date Title
Wang et al. Deep learning method based on gated recurrent unit and variational mode decomposition for short-term wind power interval prediction
Jiang et al. The two-stage machine learning ensemble models for stock price prediction by combining mode decomposition, extreme learning machine and improved harmony search algorithm
Chen et al. Synthesizing designs with interpart dependencies using hierarchical generative adversarial networks
Zhang et al. A novel hybrid model for wind speed prediction based on VMD and neural network considering atmospheric uncertainties
Yan et al. A multiagent quantum deep reinforcement learning method for distributed frequency control of islanded microgrids
CN113193556B (en) Short-term wind power prediction method based on probability prediction model
Li et al. An intelligent transient stability assessment framework with continual learning ability
Dong et al. Surrogate-assisted teaching-learning-based optimization for high-dimensional and computationally expensive problems
CN108428021B (en) Micro-grid short-term load prediction model based on HSA-RRNN
Bubbar et al. A method for comparing wave energy converter conceptual designs based on potential power capture
CN114050607A (en) Construction system for power distribution network reconstruction digital model
Kim et al. Real-time power system transient stability prediction using convolutional layer and long short-term memory
Qiu et al. Data-driven forward-inverse problems of the 2-coupled mixed derivative nonlinear Schrödinger equation using deep learning
Nepomuceno et al. Nonlinear identification using prior knowledge of fixed points: a multiobjective approach
Prabakaran et al. Economic dispatch using hybrid particle swarm optimization with prohibited operating zones and ramp rate limit constraints
Arshad et al. Deep Deterministic Policy Gradient to Regulate Feedback Control Systems Using Reinforcement Learning.
Wu et al. Microgrid Fault Diagnosis Based on Whale Algorithm Optimizing Extreme Learning Machine
CN116978450A (en) Protein data processing method, device, electronic equipment and storage medium
Lin et al. Soft variable structure fractional sliding-mode control for frequency regulation in renewable shipboard microgrids
Peng et al. Adaptive output‐feedback quadratic tracking control of continuous‐time systems via value iteration with its application
Zheng et al. Networked synthetic dynamic PMU data generation: A generative adversarial network approach
Chaturvedi et al. Neural-Wavelet Based Hybrid Model for Short-Term Load Forecasting
Giri et al. Four-area load frequency control of an interconnected power system using neuro-fuzzy hybrid intelligent proportional and integral control approach
Huang et al. A fuzzy set based solution method for multiobjective optimal design problem of mechanical and structural systems using functional-link net
Peng et al. Multi‐rate electromagnetic transient simulation of large‐scale power system based on multi‐core

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