CN115456312A - Short-term power load prediction method and system based on octyl geometric modal decomposition - Google Patents

Short-term power load prediction method and system based on octyl geometric modal decomposition Download PDF

Info

Publication number
CN115456312A
CN115456312A CN202211395000.7A CN202211395000A CN115456312A CN 115456312 A CN115456312 A CN 115456312A CN 202211395000 A CN202211395000 A CN 202211395000A CN 115456312 A CN115456312 A CN 115456312A
Authority
CN
China
Prior art keywords
matrix
power load
dimension
value
formula
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202211395000.7A
Other languages
Chinese (zh)
Inventor
董朕
黄汉生
徐备
黄磊
吴建光
高东慧
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.)
Zhaoqing Power Supply Bureau of Guangdong Power Grid Co Ltd
Original Assignee
Zhaoqing Power Supply Bureau of Guangdong 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 Zhaoqing Power Supply Bureau of Guangdong Power Grid Co Ltd filed Critical Zhaoqing Power Supply Bureau of Guangdong Power Grid Co Ltd
Priority to CN202211395000.7A priority Critical patent/CN115456312A/en
Publication of CN115456312A publication Critical patent/CN115456312A/en
Pending legal-status Critical Current

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
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • G06N3/086—Learning methods using evolutionary algorithms, e.g. genetic algorithms or genetic programming
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • 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
    • 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
    • H02J3/003—Load forecast, e.g. methods or systems for forecasting future load demand
    • 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]
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
    • Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
    • Y04S10/00—Systems supporting electrical power generation, transmission or distribution
    • Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications

Landscapes

  • Engineering & Computer Science (AREA)
  • Business, Economics & Management (AREA)
  • Economics (AREA)
  • Health & Medical Sciences (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Human Resources & Organizations (AREA)
  • Strategic Management (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Marketing (AREA)
  • General Health & Medical Sciences (AREA)
  • Tourism & Hospitality (AREA)
  • General Business, Economics & Management (AREA)
  • Operations Research (AREA)
  • Data Mining & Analysis (AREA)
  • Power Engineering (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Game Theory and Decision Science (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Evolutionary Biology (AREA)
  • Physiology (AREA)
  • Artificial Intelligence (AREA)
  • Biomedical Technology (AREA)
  • Biophysics (AREA)
  • Computational Linguistics (AREA)
  • Quality & Reliability (AREA)
  • Evolutionary Computation (AREA)
  • Development Economics (AREA)
  • Molecular Biology (AREA)
  • Computing Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • Mathematical Physics (AREA)
  • Software Systems (AREA)
  • Public Health (AREA)
  • Water Supply & Treatment (AREA)
  • Primary Health Care (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

The invention relates to the technical field of power load processing, and discloses a short-term power load prediction method and a system based on octagon modal decomposition.

Description

Short-term power load prediction method and system based on octyl geometric modal decomposition
Technical Field
The invention relates to the technical field of power load processing, in particular to a short-term power load prediction method and system based on octyl geometric modal decomposition.
Background
The load prediction is divided into three categories, namely short-term, medium-term and long-term, wherein the short term is several minutes, and the long term can reach several months or even years. The medium and long-term load prediction is commonly used for new station operation, power grid capacity increase and reconstruction, equipment overhaul plan, reservoir optimization scheduling plan, fuel supply plan and the like. The short-term load prediction plays a crucial role in determining the optimal unit combination, reducing the rotating reserve capacity and avoiding safety accidents, and is a key component for guaranteeing the economic operation of the power grid. With the development of the power market, the power load is accurately predicted in a short period, the safe operation of a power grid can be effectively guaranteed, the power generation cost is reduced, the user requirements are met, and the social and economic benefits are improved.
In the past, power load prediction is mainly applied to a scheduling department to make planning arrangement or control strategies of power generation and power supply, research is mainly concentrated on a power generation side, and research on a single user side is rarely related, while with the arrival of a smart grid and the gradual development of distributed renewable energy, user-side power generation plays an increasingly important role in a modern grid structure, how to effectively improve the consumption of renewable energy and the efficiency of a user-side household energy management system becomes increasingly important, analysis on power utilization behaviors of the user side can more effectively promote demand-side management, reasonably inhibit load peaks and improve the utilization rate of grid assets so as to meet the arrival of big data and an intelligent era, on the other hand, a power load prediction method has a certain universality for regional loads and personal loads, compared with regional loads, an individual user load generally has stronger random characteristics, a model with stronger high-frequency random characteristic processing capability can obtain a better effect, and a model capable of predicting the individual user load generally can obtain a more ideal result for regional load prediction with a smoother rule.
Because the load of the power system has a certain periodic characteristic and the factors influencing the load are complex (weather, economy, holidays, observation errors and the like), the load of the power system presents stronger randomness and non-periodic components, and great difficulty is brought to short-term prediction.
At present, short-term load prediction methods can be divided into traditional statistical prediction methods and machine learning prediction methods, statistics comprise generalized autoregressive conditional variance, time series and the like, and load prediction is realized by learning load recursion relations at different moments. With the wide application of machine learning, many researchers apply it to load prediction. Such as a neural Network, an extreme learning machine, a long-time and short-time memory Network, an Echo State Network (ESN), etc., the load sequence has a certain time sequence correlation with the input characteristics of the load prediction model, and if the prediction model can learn the hidden relationship, the prediction accuracy of the short-time load is improved.
However, the ESN parameters are selected only by experience, so that the uncertainty is very high, the prediction accuracy is poor, meanwhile, the basic SMA algorithm is prone to the problems of unstable optimization result, low convergence speed, local optimization and the like when a high-dimensional complex function and a function with an optimal solution not at the origin are optimized, and in addition, the randomness and non-periodic components of the power load on the user side are difficult to process by a single prediction method, so that the modal aliasing problem exists, and the prediction accuracy of the power load is influenced.
Disclosure of Invention
The invention provides a short-term power load prediction method and system based on octyl geometric modal decomposition, and solves the technical problem of poor prediction accuracy of power loads.
In view of this, the first aspect of the present invention provides a short-term power load prediction method based on simmer geometry decomposition, including the following steps:
s1, obtaining historical power loads of a user side, and constructing a historical power load time sequence;
s2, decomposing the historical power load time sequence by adopting a sine geometric mode to obtain a plurality of corresponding load components;
s3, dividing all load components into a training data set and a testing data set;
s4, constructing an echo state network, and optimizing the echo state network through a dimension competition myxomycete algorithm to obtain an initial power load prediction model;
s5, inputting the load components in the training data set to the initial power load prediction model one by one for training to obtain power load prediction submodels corresponding to the load components;
s6, inputting the test data set into the power load forecasting submodel for forecasting to obtain a corresponding load forecasting value;
and S7, performing superposition processing on all the load predicted values to obtain a final user side power load prediction result.
Preferably, step S2 specifically includes:
s201, setting a historical power load time sequence as
Figure 889055DEST_PATH_IMAGE001
In which
Figure 51046DEST_PATH_IMAGE002
For the data length, the time sequence of the power load is reconstructed by adopting a time sequence delay topological equivalent method to obtain a track matrix
Figure 556284DEST_PATH_IMAGE003
Comprises the following steps:
Figure 778318DEST_PATH_IMAGE004
in the formula,
Figure 190845DEST_PATH_IMAGE005
in order to embed the dimension number of the dimension,
Figure 140215DEST_PATH_IMAGE006
in order to delay the time of the delay,
Figure 321798DEST_PATH_IMAGE007
；
s202, performing autocorrelation analysis on the track matrix X to obtain a covariance symmetric matrix
Figure 714733DEST_PATH_IMAGE008
Comprises the following steps:
Figure 801507DEST_PATH_IMAGE009
in the formula, T is a transposed symbol;
s203, a covariance symmetric matrix is matched
Figure 367617DEST_PATH_IMAGE008
Constructing a Hamilton matrix
Figure 75810DEST_PATH_IMAGE010
Comprises the following steps:
Figure 154493DEST_PATH_IMAGE011
s204, aiming at the Hamilton matrix
Figure 541612DEST_PATH_IMAGE010
Squaring to obtain another Hamilton matrix
Figure 583518DEST_PATH_IMAGE012
Comprises the following steps:
Figure 739693DEST_PATH_IMAGE013
s205, constructing a sine orthogonal matrix through the following formula
Figure 726628DEST_PATH_IMAGE014
For Hamilton matrix
Figure 335464DEST_PATH_IMAGE012
And (3) carrying out octyl orthogonal transformation:
Figure 446639DEST_PATH_IMAGE015
in the formula,
Figure 644271DEST_PATH_IMAGE016
in the form of a real matrix,
Figure 612227DEST_PATH_IMAGE017
for upper triangular matrices, the matrix elements of upper triangular matrices
Figure 646042DEST_PATH_IMAGE018
Wherein i and j are respectively the number of rows and columns of the matrix
Figure 279018DEST_PATH_IMAGE008
、
Figure 409785DEST_PATH_IMAGE017
And
Figure 220746DEST_PATH_IMAGE003
is expressed as:
Figure 256704DEST_PATH_IMAGE019
；
Figure 506420DEST_PATH_IMAGE020
is a matrix of a Householder, and the matrix of the Householder,
Figure 429376DEST_PATH_IMAGE021
wherein
Figure 739135DEST_PATH_IMAGE022
，
Figure 999739DEST_PATH_IMAGE023
；
s206, calculating an upper triangular matrix by using QR decomposition
Figure 990829DEST_PATH_IMAGE017
Characteristic value of (D) is noted
Figure 830609DEST_PATH_IMAGE024
From the upper triangular matrix
Figure 498220DEST_PATH_IMAGE017
Eigenvalue calculation matrix of
Figure 790661DEST_PATH_IMAGE008
The characteristic values of (A) are:
Figure 585442DEST_PATH_IMAGE025
for matrix
Figure 466679DEST_PATH_IMAGE008
The characteristic value of the vector is vectorized to obtain a corresponding characteristic vector which is recorded as
Figure 118240DEST_PATH_IMAGE026
；
S207, passing matrix
Figure 835660DEST_PATH_IMAGE008
Feature vector of
Figure 230870DEST_PATH_IMAGE026
And calculating the sum track matrix X to obtain a conversion coefficient matrix
Figure 232192DEST_PATH_IMAGE027
Comprises the following steps:
Figure 320234DEST_PATH_IMAGE028
using feature vector matrices
Figure 524951DEST_PATH_IMAGE020
And a conversion coefficient matrix
Figure 648152DEST_PATH_IMAGE027
Calculating to obtain a reconstruction matrix
Figure 317031DEST_PATH_IMAGE029
：
Figure 513657DEST_PATH_IMAGE030
Figure 2407DEST_PATH_IMAGE029
Representing an initial one-component matrix of the image,
Figure 191949DEST_PATH_IMAGE031
，
Figure 715334DEST_PATH_IMAGE032
wherein the matrix is reconstructedZIs composed of
Figure 817282DEST_PATH_IMAGE033
Matrix of
Figure 245858DEST_PATH_IMAGE005
An initial single component matrix
Figure 786561DEST_PATH_IMAGE029
Is prepared by the following steps of;
s208, for any initial single-component matrix
Figure 836557DEST_PATH_IMAGE029
Defining each element of the matrix as
Figure 889832DEST_PATH_IMAGE034
In which
Figure 353175DEST_PATH_IMAGE035
Let us order
Figure 635251DEST_PATH_IMAGE036
And, and
Figure 133229DEST_PATH_IMAGE037
wherein, in the process,
Figure 653396DEST_PATH_IMAGE038
representing a single component matrix
Figure 807296DEST_PATH_IMAGE029
The minimum of the row and the column of (c),
Figure 689802DEST_PATH_IMAGE039
representing a single component matrix
Figure 229236DEST_PATH_IMAGE029
Of rows and columns, wherein when
Figure 843888DEST_PATH_IMAGE040
When it is taken
Figure 281823DEST_PATH_IMAGE041
When is coming into contact with
Figure 154970DEST_PATH_IMAGE042
Taking out
Figure 361960DEST_PATH_IMAGE043
Wherein, in the process,
Figure 147514DEST_PATH_IMAGE044
the diagonal average transformation matrix calculation element values are represented, and the diagonal average transformation matrix is represented as follows:
Figure 72744DEST_PATH_IMAGE045
in the formula,
Figure 749582DEST_PATH_IMAGE046
representing the diagonal average transformation matrix calculation element values,
Figure 14342DEST_PATH_IMAGE047
and
Figure 298692DEST_PATH_IMAGE048
respectively representing the values of the elements of the diagonal averaging transformation matrix calculation
Figure 635520DEST_PATH_IMAGE047
And row and column
Figure 663519DEST_PATH_IMAGE048
The columns of the image data are,
Figure 782785DEST_PATH_IMAGE049
representing the kth element value of the diagonal average transformation matrix;
s209, converting the initial single component matrix by diagonal averaging
Figure 238037DEST_PATH_IMAGE029
Conversion to a one-dimensional time series of single-component signals, denoted
Figure 324811DEST_PATH_IMAGE050
，
Figure 828604DEST_PATH_IMAGE051
Representing the data length, i is more than or equal to 1 and less than or equal to d, and sequentially performing matrix comparison on all initial single components
Figure 864693DEST_PATH_IMAGE052
Is subjected to diagonal averaging to obtain
Figure 677797DEST_PATH_IMAGE005
A one-dimensional time series, pair
Figure 799337DEST_PATH_IMAGE005
And superposing the one-dimensional time sequences to obtain a single-component total signal matrix as follows:
Figure 841243DEST_PATH_IMAGE053
wherein the single component total signal matrixYIs one
Figure 997417DEST_PATH_IMAGE054
A matrix;
s210, initial single component
Figure 247002DEST_PATH_IMAGE055
Form a group in sequence
Figure 793521DEST_PATH_IMAGE056
Vector of vitamin the vector is a vector of a number of,
Figure 701434DEST_PATH_IMAGE057
Figure 901996DEST_PATH_IMAGE058
representing the original single component
Figure 869952DEST_PATH_IMAGE055
To middle
Figure 903767DEST_PATH_IMAGE059
The dots start to be continuous
Figure 615371DEST_PATH_IMAGE056
An
Figure 933089DEST_PATH_IMAGE060
A value of (d);
s211, defining vector
Figure 9629DEST_PATH_IMAGE058
And
Figure 514429DEST_PATH_IMAGE061
in betweenDistance between two adjacent devices
Figure 701828DEST_PATH_IMAGE062
The element with the largest difference is the corresponding element:
Figure 139631DEST_PATH_IMAGE063
in the formula,
Figure 387073DEST_PATH_IMAGE061
representing the original single component
Figure 192218DEST_PATH_IMAGE055
The j point in the figure begins continuously
Figure 435505DEST_PATH_IMAGE056
An
Figure 275285DEST_PATH_IMAGE060
The value of (a) is set to (b),
Figure 693628DEST_PATH_IMAGE064
；
Figure 438599DEST_PATH_IMAGE065
，
Figure 30117DEST_PATH_IMAGE066
；
s212, mixing
Figure 927666DEST_PATH_IMAGE062
Comparing with a preset threshold value, wherein the preset threshold value is 0.1 to 0.25
Figure 579227DEST_PATH_IMAGE067
Wherein, in the process,
Figure 545915DEST_PATH_IMAGE067
determining standard deviation for single component signal
Figure 941124DEST_PATH_IMAGE062
The number of the distance is less than the preset threshold value, and the sum of the number and the distance is
Figure 427600DEST_PATH_IMAGE068
The ratio was calculated as:
Figure 250063DEST_PATH_IMAGE069
to find out their pairs
Figure 969626DEST_PATH_IMAGE056
Vector of vitamin of vectors average values are:
Figure 168526DEST_PATH_IMAGE070
s213, dimension of
Figure 775088DEST_PATH_IMAGE056
Plus 1, the sequence of the serial number arrangement forms a new group
Figure 946613DEST_PATH_IMAGE056
The + 1-dimensional vector is,
Figure 700943DEST_PATH_IMAGE071
Figure 641217DEST_PATH_IMAGE072
represents from the first
Figure 164602DEST_PATH_IMAGE059
The points beginning to be continuous
Figure 515818DEST_PATH_IMAGE056
+1 pieces
Figure 757443DEST_PATH_IMAGE060
A value of (d);
s214, defining a vector
Figure 235829DEST_PATH_IMAGE072
And
Figure 613721DEST_PATH_IMAGE073
distance between them
Figure 401417DEST_PATH_IMAGE074
The element with the largest difference is the one with the largest difference between the two corresponding elements, namely:
Figure 536864DEST_PATH_IMAGE075
in the formula,
Figure 881257DEST_PATH_IMAGE073
indicating continuation from the j-th point
Figure 566185DEST_PATH_IMAGE056
+1 pieces
Figure 337832DEST_PATH_IMAGE060
The value of (a) is,
Figure 226154DEST_PATH_IMAGE076
；
Figure 108659DEST_PATH_IMAGE065
，
Figure 651024DEST_PATH_IMAGE066
；
s215, mixing
Figure 531255DEST_PATH_IMAGE077
Comparing with a preset threshold value to determine
Figure 969189DEST_PATH_IMAGE077
The number of the distance is less than the preset threshold value, and the sum of the number and the distance is
Figure 842336DEST_PATH_IMAGE068
The ratio was calculated as:
Figure 49327DEST_PATH_IMAGE078
find it to
Figure 100459DEST_PATH_IMAGE056
The average value of the +1 vector is:
Figure 25690DEST_PATH_IMAGE079
the formula for calculating the sample entropy value is as follows:
Figure 436949DEST_PATH_IMAGE080
s216, differentiating the entropy values of the samples
Figure 436129DEST_PATH_IMAGE081
Initial single component of
Figure 454900DEST_PATH_IMAGE055
Adding to obtain new components
Figure 319957DEST_PATH_IMAGE082
Namely:
Figure 347956DEST_PATH_IMAGE083
in the formula,
Figure 467222DEST_PATH_IMAGE084
the component obtained by the octyl geometric modal decomposition is represented, and a represents the a-th component sequence.
Preferably, step S4 specifically includes:
s401, establishing an echo state network as follows:
Figure 922474DEST_PATH_IMAGE085
Figure 746598DEST_PATH_IMAGE086
Figure 312708DEST_PATH_IMAGE087
wherein,
Figure 552060DEST_PATH_IMAGE088
in order to input the dimension number, the dimension number is input,
Figure 365164DEST_PATH_IMAGE089
the number of internal neurons, l is the output dimension, and u (v), x (v) and y (v) are the input vector, the state vector and the output vector of the echo state network respectively;
s402, according to the input vector, the state vector and the output vector of the echo state network, training the echo state network by the following formula to obtain:
Figure 486704DEST_PATH_IMAGE090
Figure 794188DEST_PATH_IMAGE091
wherein f () is an internal neuron activation function Sigmoid, f out () As a function of the output layer(s),Wa connection weight matrix for internal states to internal states,
Figure 950363DEST_PATH_IMAGE092
for randomly generated input layers to a reserve pool
Figure 934368DEST_PATH_IMAGE093
×
Figure 215308DEST_PATH_IMAGE094
The order of the connection weight matrix is,
Figure 123221DEST_PATH_IMAGE095
feeding back to reserve tank for randomly generated output layers
Figure 586433DEST_PATH_IMAGE093
A connection weight matrix of order x 1,
Figure 554389DEST_PATH_IMAGE096
is from reserve pool to output layer
Figure 588204DEST_PATH_IMAGE093
+
Figure 299808DEST_PATH_IMAGE094
+ l) order output weight matrix;
s403, optimizing the echo state network through a dimension competition slime algorithm, which specifically comprises the following steps:
1) Setting the total number U of individual slime organisms and the maximum iteration number of the dimension competition slime organism algorithm
Figure 354876DEST_PATH_IMAGE097
Proportional parameter z, decreasing parameter
Figure 431416DEST_PATH_IMAGE098
Random number
Figure 14844DEST_PATH_IMAGE099
Determining parameters of the updating method of the slime location
Figure 717090DEST_PATH_IMAGE100
Dimension D of individual slime bacteria, dimension competition probability Pv, and Gaussian variation probability
Figure 436784DEST_PATH_IMAGE101
；
2) Randomly generating a group of solutions as initial parameters to fit a dimension competition myxomycete algorithm to optimize an echo state network:
Figure 684226DEST_PATH_IMAGE102
wherein i =1,2.., U;
Figure 754950DEST_PATH_IMAGE103
in order to store the scale of the neurons in the pool,
Figure 995308DEST_PATH_IMAGE104
in order to be the radius of the spectrum,
Figure 507192DEST_PATH_IMAGE105
in order to be sparse in degree,
Figure 987852DEST_PATH_IMAGE106
in order to input the dimensions of the cell,
Figure 998402DEST_PATH_IMAGE107
in order to displace the input unit, the displacement of the input unit,
Figure 589920DEST_PATH_IMAGE108
in order to output the dimensions of the cell,
Figure 221890DEST_PATH_IMAGE109
displacing the output unit;
3) Virtually exploring a target space through initial parameters, and in t +1 iterations of the target space, updating the positions of the slime individuals in the following ways:
Figure 139030DEST_PATH_IMAGE110
Figure 120366DEST_PATH_IMAGE111
in the formula,
Figure 453259DEST_PATH_IMAGE112
and
Figure 267631DEST_PATH_IMAGE113
in order to search the upper and lower boundaries of the range,
Figure 277044DEST_PATH_IMAGE114
is composed of
Figure 544077DEST_PATH_IMAGE115
Z is a ratio parameter for determining the ratio of randomly distributed slime bacteria individuals to slime bacteria,
Figure 680661DEST_PATH_IMAGE116
the position of the highest food odor concentration currently found, namely the optimal solution position;
Figure 349539DEST_PATH_IMAGE003
the current position of the slime mold;
Figure 529854DEST_PATH_IMAGE117
and
Figure 956287DEST_PATH_IMAGE116
the positions of two individuals randomly selected from the group are respectively;
Figure 693299DEST_PATH_IMAGE118
is the current iteration number;
Figure 669214DEST_PATH_IMAGE119
is a coefficient, the value of which is
Figure 99059DEST_PATH_IMAGE120
And gradually approaches 0 as the number of iterations increases, wherein,
Figure 278367DEST_PATH_IMAGE121
，
Figure 743371DEST_PATH_IMAGE097
is the maximum iteration number;
Figure 386842DEST_PATH_IMAGE098
for a decreasing parameter from 1 to 0,
Figure 659691DEST_PATH_IMAGE099
is composed of
Figure 388613DEST_PATH_IMAGE115
A random number in between, and a random number,
Figure 654378DEST_PATH_IMAGE100
in order to determine the parameters of the updating method of the slime location,
Figure 90039DEST_PATH_IMAGE122
，
Figure 861686DEST_PATH_IMAGE123
is shown as
Figure 999275DEST_PATH_IMAGE059
The fitness value of each individual slime mold,
Figure 881780DEST_PATH_IMAGE124
representing the optimal fitness value of the slime mold under the current iteration times;
the fitness value calculation formula is as follows:
Figure 906368DEST_PATH_IMAGE125
in the formula,
Figure 848916DEST_PATH_IMAGE126
、
Figure 739381DEST_PATH_IMAGE127
respectively predicting an actual value of the user side short-term power load and a predicted output value of the user side short-term power load; g is the number of training samples;
slime self-adaptive weight
Figure 97681DEST_PATH_IMAGE128
The expression of (a) is:
Figure 570250DEST_PATH_IMAGE129
in the formula,
Figure 873580DEST_PATH_IMAGE099
is composed of
Figure 798811DEST_PATH_IMAGE115
A random number in between, and a random number,
Figure 960802DEST_PATH_IMAGE130
indicating the individual position indexes after the fitness values are arranged in ascending order,
Figure 22299DEST_PATH_IMAGE131
represent
Figure 493600DEST_PATH_IMAGE132
The population of the first half of the middle rank,
Figure 906127DEST_PATH_IMAGE133
represents the optimal fitness value obtained by the current iterative process,
Figure 871809DEST_PATH_IMAGE134
representing the worst fitness value obtained by the current iteration process;
4) To slime bacteria individual
Figure 240343DEST_PATH_IMAGE135
All dimensions are randomly paired without repeating pairwise pairs, the total is D/2 pairs, and any pair of dimensions is paired, if rand<Pv, performing a dimension crossover operator on the pair of dimensions according to the following formula;
Figure 430015DEST_PATH_IMAGE136
Figure 1942DEST_PATH_IMAGE137
；
Figure 568053DEST_PATH_IMAGE138
in the formula,
Figure 322251DEST_PATH_IMAGE139
is a slime mold individual
Figure 682825DEST_PATH_IMAGE135
To (1) a
Figure 7627DEST_PATH_IMAGE140
And the first
Figure 377429DEST_PATH_IMAGE141
Dimension generation by dimension crossing;
Figure 457905DEST_PATH_IMAGE142
is [0,1 ]]A random number in between, and a random number,
Figure 192643DEST_PATH_IMAGE143
is the dimension cross probability;
5) Calculating progeny according to the formula
Figure 801478DEST_PATH_IMAGE144
With parent Myxomycetes
Figure 161922DEST_PATH_IMAGE135
And updating the individual positions of the slime bacteria, and recording the current global optimal solution
Figure 172603DEST_PATH_IMAGE145
；
Figure 78242DEST_PATH_IMAGE146
6) If it is
Figure 174374DEST_PATH_IMAGE147
，
Figure 72929DEST_PATH_IMAGE148
Then entering a Gaussian mutation operator to carry out optimization on the optimal individuals
Figure 938116DEST_PATH_IMAGE145
Performing Gaussian mutation operation, further performing local search, and updating
Figure 749078DEST_PATH_IMAGE145
The position of (2):
Figure 785036DEST_PATH_IMAGE149
Figure 34751DEST_PATH_IMAGE150
in the formula,
Figure 957708DEST_PATH_IMAGE151
the particles are the optimal particles after Gaussian variation, N (0, 1) is a Gaussian distribution random quantity with the mean value of 0 and the variance of 1;
7) Judging the current iteration number
Figure 267467DEST_PATH_IMAGE118
Whether or not the maximum number of iterations has been reached
Figure 516352DEST_PATH_IMAGE097
If so, the iteration is finished, the optimal solution is output, otherwise,
Figure 507442DEST_PATH_IMAGE118
+1, returning to the step 2) to continue searching until the current iteration times
Figure 347222DEST_PATH_IMAGE118
To maximum number of iterations
Figure 14833DEST_PATH_IMAGE097
And after iteration stops, outputting a current global optimal solution, and updating initial parameters of the echo state network by using the global optimal solution to obtain an initial power load prediction model.
In a second aspect, the present invention further provides a short-term power load prediction system based on simmer geometry decomposition, including:
the load acquisition module is used for acquiring historical power loads of a user side and constructing a historical power load time sequence;
the decomposition module is used for decomposing the historical power load time sequence by adopting a sine-shaped geometric mode to obtain a plurality of corresponding load components;
the dividing module is used for dividing all load components into a training data set and a testing data set;
the network construction module is used for constructing an echo state network, and optimizing the echo state network through a dimension competition myxobacteria algorithm to obtain an initial power load prediction model;
the training module is used for inputting the load components in the training data set to the initial power load prediction model one by one for training to obtain a power load prediction submodel corresponding to each load component;
the prediction module is used for inputting the test data set into the power load prediction submodel for prediction to obtain a corresponding load prediction value;
and the superposition module is used for carrying out superposition processing on all the load predicted values to obtain a final user side power load prediction result.
According to the technical scheme, the invention has the following advantages:
according to the method, the historical power load time sequence is decomposed by adopting a octave geometric mode to obtain a plurality of corresponding load components, the influence of the volatility of the user side power load time sequence on a prediction result is reduced, a prediction model of a dimension competition slime algorithm optimized echo state network is respectively established for each component, the stability and the generalization capability of the prediction model are improved, the prediction values of all the components are superposed to obtain an actual user side power load prediction result, and the prediction accuracy of the power load is improved.
Drawings
Fig. 1 is a flowchart of a short-term power load prediction method based on simmer geometric mode decomposition according to an embodiment of the present invention;
fig. 2 is a schematic structural diagram of a short-term power load prediction system based on simmer geometry mode decomposition according to an embodiment of the present invention.
Detailed Description
In order to make the technical solutions of the present invention better understood, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of 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.
For easy understanding, referring to fig. 1, the method for predicting short-term power load based on simmer geometry mode decomposition according to the present invention includes the following steps:
s1, obtaining historical power loads of a user side, and constructing a historical power load time sequence;
s2, decomposing the historical power load time sequence by adopting a sine geometric mode to obtain a plurality of corresponding load components;
s3, dividing all load components into a training data set and a testing data set;
s4, constructing an echo state network, and optimizing the echo state network through a dimension competition myxomycete algorithm to obtain an initial power load prediction model;
s5, inputting the load components in the training data set to the initial power load prediction model one by one for training to obtain a power load prediction submodel corresponding to each load component;
s6, inputting the test data set into a power load forecasting sub-model for forecasting to obtain a corresponding load forecasting value;
and S7, performing superposition processing on all the load predicted values to obtain a final user side power load prediction result.
It can be understood that each load predicted value is a predicted result of the decomposed component, and the predicted values of all the components need to be superimposed to obtain a final load predicted value.
It should be noted that, this embodiment provides a short-term power load prediction method based on simmerical geometry mode decomposition, which decomposes a historical power load time sequence by using a simmerical geometry mode to obtain a plurality of corresponding load components, reduces the influence of the volatility of the user-side power load time sequence on a prediction result, establishes a prediction model for optimizing an echo state network by using a dimension competition slime mold algorithm for each component, improves the stability and generalization capability of the prediction model, and superimposes the prediction values of all the components to obtain an actual user-side power load prediction result, thereby improving the prediction accuracy of the power load.
In a specific embodiment, step S2 specifically includes:
s201, setting a historical power load time sequence as
Figure 307274DEST_PATH_IMAGE001
Wherein
Figure 102055DEST_PATH_IMAGE002
For the data length, the time sequence of the power load is reconstructed by adopting a time sequence delay topological equivalent method to obtain a track matrix
Figure 796341DEST_PATH_IMAGE003
Comprises the following steps:
Figure 634853DEST_PATH_IMAGE004
in the formula,
Figure 680169DEST_PATH_IMAGE005
for embedding dimensionThe number of the first and second groups is,
Figure 747483DEST_PATH_IMAGE006
in order to delay the time of the process,
Figure 748805DEST_PATH_IMAGE007
；
wherein a suitable embedding dimension is selected
Figure 836847DEST_PATH_IMAGE005
And a delay time
Figure 41564DEST_PATH_IMAGE006
The corresponding reconstruction matrix is obtained
Figure 240464DEST_PATH_IMAGE003
. Since different embedding dimensions can produce different effects, the idea of determining the embedding dimensions is adopted to calculate the initial time series
Figure 833644DEST_PATH_IMAGE152
The frequency corresponding to the maximum peak value of the PSD is the Power Spectral Density (PSD) of
Figure 30270DEST_PATH_IMAGE153
. If the normalized frequency is less than a given threshold
Figure 971550DEST_PATH_IMAGE154
Then is provided with
Figure 708562DEST_PATH_IMAGE005
Is composed of
Figure 169630DEST_PATH_IMAGE155
Wherein
Figure 599474DEST_PATH_IMAGE002
Is the data length. Otherwise the embedding dimension is set to
Figure 762471DEST_PATH_IMAGE156
，
Figure 975278DEST_PATH_IMAGE157
Is the sampling frequency. Delay time
Figure 618749DEST_PATH_IMAGE006
Is usually taken
Figure 406445DEST_PATH_IMAGE158
。
S202, performing autocorrelation analysis on the track matrix X to obtain a covariance symmetric matrix
Figure 135367DEST_PATH_IMAGE008
Comprises the following steps:
Figure 151864DEST_PATH_IMAGE009
in the formula, T is a transposed symbol;
s203, a covariance symmetric matrix is matched
Figure 839722DEST_PATH_IMAGE008
Constructing a Hamilton matrix
Figure 345790DEST_PATH_IMAGE010
Comprises the following steps:
Figure 499691DEST_PATH_IMAGE011
s204, aiming at Hamilton matrix
Figure 569147DEST_PATH_IMAGE010
Squaring to obtain another Hamilton matrix
Figure 921631DEST_PATH_IMAGE012
Comprises the following steps:
Figure 801862DEST_PATH_IMAGE013
S205constructing a sinc-orthogonal matrix by the following formula
Figure 239796DEST_PATH_IMAGE014
For Hamilton matrix
Figure 847364DEST_PATH_IMAGE012
And (3) carrying out octyl orthogonal transformation:
Figure 992038DEST_PATH_IMAGE015
in the formula,
Figure 839908DEST_PATH_IMAGE016
is a real matrix and is characterized in that,
Figure 217669DEST_PATH_IMAGE017
for upper triangular matrices, the matrix elements of an upper triangular matrix
Figure 379660DEST_PATH_IMAGE018
Wherein i and j are respectively the number of rows and columns of the matrix
Figure 706736DEST_PATH_IMAGE008
、
Figure 180967DEST_PATH_IMAGE017
And
Figure 531177DEST_PATH_IMAGE003
is expressed as:
Figure 293597DEST_PATH_IMAGE019
；
Figure 662130DEST_PATH_IMAGE020
in the form of a Householder matrix,
Figure 55065DEST_PATH_IMAGE021
wherein, in the process,
Figure 954888DEST_PATH_IMAGE022
，
Figure 707949DEST_PATH_IMAGE023
；
wherein, the sine orthogonal matrixHHas the property of sine matrix, so that Hamilton matrix
Figure 744038DEST_PATH_IMAGE012
The structure of (2) is not destroyed in the matrix transformation process.
Meanwhile, based on the mathematical theorem: suppose that
Figure 42296DEST_PATH_IMAGE159
Is a matrix of tracks, an
Figure 881944DEST_PATH_IMAGE009
Is a symmetric matrix. By matrix
Figure 986167DEST_PATH_IMAGE008
Constructing a Hamiltonian matrix
Figure 80025DEST_PATH_IMAGE010
I.e. by
Figure 877079DEST_PATH_IMAGE011
。
Having a Householder matrixHThen it can pass through
Figure 687514DEST_PATH_IMAGE160
Transformation structure upper triangular Hessenberg matrix
Figure 798690DEST_PATH_IMAGE017
Namely:
Figure 809371DEST_PATH_IMAGE161
Figure 964278DEST_PATH_IMAGE162
Figure 60410DEST_PATH_IMAGE163
s206, calculating an upper triangular matrix by QR decomposition
Figure 444118DEST_PATH_IMAGE017
Characteristic value of (D) is recorded as
Figure 761836DEST_PATH_IMAGE024
From the upper triangular matrix
Figure 900693DEST_PATH_IMAGE017
Eigenvalue calculation matrix of
Figure 156225DEST_PATH_IMAGE008
The characteristic values of (A) are:
Figure 405941DEST_PATH_IMAGE025
for matrix
Figure 843744DEST_PATH_IMAGE008
The characteristic value of the vector is vectorized to obtain a corresponding characteristic vector which is recorded as
Figure 91186DEST_PATH_IMAGE026
；
S207, passing matrix
Figure 161910DEST_PATH_IMAGE008
Feature vector of
Figure 405197DEST_PATH_IMAGE026
And calculating the sum track matrix X to obtain a conversion coefficient matrix
Figure 244977DEST_PATH_IMAGE027
Comprises the following steps:
Figure 663320DEST_PATH_IMAGE028
using eigenvector matrices
Figure 955761DEST_PATH_IMAGE020
And a conversion coefficient matrix
Figure 999809DEST_PATH_IMAGE027
Calculating to obtain a reconstruction matrix
Figure 694096DEST_PATH_IMAGE029
：
Figure 283340DEST_PATH_IMAGE030
Figure 515607DEST_PATH_IMAGE029
The initial single-component matrix is represented,
Figure 910816DEST_PATH_IMAGE031
，
Figure 131713DEST_PATH_IMAGE032
wherein the matrix is reconstructedZIs composed of
Figure 219755DEST_PATH_IMAGE033
Matrix of
Figure 939318DEST_PATH_IMAGE005
An initial single component matrix
Figure 810322DEST_PATH_IMAGE029
Is prepared from (A) and (B);
s208, for any initial single-component matrix
Figure 479201DEST_PATH_IMAGE029
Defining each element of the matrix as
Figure 928024DEST_PATH_IMAGE034
Wherein
Figure 416775DEST_PATH_IMAGE035
Let us order
Figure 357049DEST_PATH_IMAGE036
And an
Figure 67385DEST_PATH_IMAGE037
Wherein, in the process,
Figure 231650DEST_PATH_IMAGE038
representing a single component matrix
Figure 410958DEST_PATH_IMAGE029
The minimum of the row and the column of (c),
Figure 951661DEST_PATH_IMAGE039
representing a single component matrix
Figure 516504DEST_PATH_IMAGE029
Is maximum of row and column, wherein
Figure 789353DEST_PATH_IMAGE040
When it is taken
Figure 252695DEST_PATH_IMAGE041
When is coming into contact with
Figure 49619DEST_PATH_IMAGE042
Get it
Figure 547596DEST_PATH_IMAGE043
Wherein
Figure 991347DEST_PATH_IMAGE044
representing calculated values of elements of a diagonal averaging transformation matrix, diagonal averagingThe equalization transformation matrix is represented as follows:
Figure 397445DEST_PATH_IMAGE045
in the formula,
Figure 279951DEST_PATH_IMAGE046
representing the diagonal average transformation matrix calculation element values,
Figure 570118DEST_PATH_IMAGE047
and
Figure 247087DEST_PATH_IMAGE048
respectively representing the values of the elements of the diagonal averaging transformation matrix calculation
Figure 871972DEST_PATH_IMAGE047
And row and column
Figure 495851DEST_PATH_IMAGE048
The columns of the image data are,
Figure 702842DEST_PATH_IMAGE049
representing the kth element value of a diagonal averaging transformation matrix;
s209, converting the initial single component matrix by diagonal averaging
Figure 3242DEST_PATH_IMAGE029
Conversion to a one-dimensional time series of single-component signals, denoted
Figure 928473DEST_PATH_IMAGE050
，
Figure 824884DEST_PATH_IMAGE051
Representing the data length, i is more than or equal to 1 and less than or equal to d, and sequentially performing matrix comparison on all initial single components
Figure 151961DEST_PATH_IMAGE052
Is obtained by carrying out diagonal averaging
Figure 888841DEST_PATH_IMAGE005
A one-dimensional time series, pair
Figure 973472DEST_PATH_IMAGE005
Superposing the one-dimensional time sequences to obtain a single-component total signal matrix as follows:
Figure 1471DEST_PATH_IMAGE053
wherein the single component total signal matrixYIs one
Figure 361215DEST_PATH_IMAGE054
A matrix;
s210, initial single component
Figure 816467DEST_PATH_IMAGE055
Form a group in sequence
Figure 653973DEST_PATH_IMAGE056
The vector of the vectors is,
Figure 407034DEST_PATH_IMAGE057
Figure 443124DEST_PATH_IMAGE058
representing the original single component
Figure 6960DEST_PATH_IMAGE055
To middle
Figure 128500DEST_PATH_IMAGE059
The dots start to be continuous
Figure 685252DEST_PATH_IMAGE056
An
Figure 513531DEST_PATH_IMAGE060
A value of (d);
it should be noted that, unlike the periodic similarity comparison and reconstruction method of the conventional octyl geometric modal decomposition, the present embodiment introduces the sample entropy into the reconstruction process of the single-component total signal, and aims to adaptively reconstruct the single-component total signal and improve the accuracy of the analysis result.
S211, defining vector
Figure 310585DEST_PATH_IMAGE058
And
Figure 371951DEST_PATH_IMAGE061
distance between them
Figure 279864DEST_PATH_IMAGE062
The element with the largest difference is the one with the largest difference between the two corresponding elements, namely:
Figure 228229DEST_PATH_IMAGE063
in the formula,
Figure 386065DEST_PATH_IMAGE061
representing a single component from the origin
Figure 419880DEST_PATH_IMAGE055
The j point of the medium is continuous
Figure 318435DEST_PATH_IMAGE056
An
Figure 449202DEST_PATH_IMAGE060
The value of (a) is,
Figure 525742DEST_PATH_IMAGE064
；
Figure 843591DEST_PATH_IMAGE065
，
Figure 545837DEST_PATH_IMAGE066
；
s212, mixing
Figure 203214DEST_PATH_IMAGE062
Comparing with a preset threshold value, wherein the preset threshold value is 0.1 to 0.25
Figure 512973DEST_PATH_IMAGE067
Wherein
Figure 770648DEST_PATH_IMAGE067
determining standard deviation for single component signal
Figure 761738DEST_PATH_IMAGE062
The number of the distance is less than the preset threshold value, and the sum of the number and the distance is
Figure 601518DEST_PATH_IMAGE068
The ratio was calculated as:
Figure 272058DEST_PATH_IMAGE069
to find out their pairs
Figure 830078DEST_PATH_IMAGE056
Of vector of dimension average values are:
Figure 359280DEST_PATH_IMAGE070
s213. Dimension
Figure 506096DEST_PATH_IMAGE056
Plus 1, the sequence of the serial number arrangement forms a new group
Figure 157657DEST_PATH_IMAGE056
The + 1-dimensional vector is,
Figure 875078DEST_PATH_IMAGE071
Figure 270287DEST_PATH_IMAGE072
represents from the first
Figure 271610DEST_PATH_IMAGE059
The points beginning to be continuous
Figure 94072DEST_PATH_IMAGE056
+1 pieces
Figure 33209DEST_PATH_IMAGE060
A value of (d);
s214, defining vector
Figure 684639DEST_PATH_IMAGE072
And
Figure 353518DEST_PATH_IMAGE073
distance between them
Figure 284565DEST_PATH_IMAGE074
The element with the largest difference is the one with the largest difference between the two corresponding elements, namely:
Figure 38894DEST_PATH_IMAGE075
in the formula,
Figure 231366DEST_PATH_IMAGE073
indicating continuation from the j-th point
Figure 754751DEST_PATH_IMAGE056
+1 pieces
Figure 856699DEST_PATH_IMAGE060
The value of (a) is,
Figure 285276DEST_PATH_IMAGE076
；
Figure 825978DEST_PATH_IMAGE065
，
Figure 141553DEST_PATH_IMAGE066
；
s215, mixing
Figure 742299DEST_PATH_IMAGE077
Comparing with a preset threshold value to determine
Figure 392592DEST_PATH_IMAGE077
The number of the distance is less than the preset threshold value, and the sum of the number and the distance is
Figure 736985DEST_PATH_IMAGE068
The ratio was calculated as:
Figure 641488DEST_PATH_IMAGE078
find it to
Figure 865664DEST_PATH_IMAGE056
The average value of the +1 vector is:
Figure 816303DEST_PATH_IMAGE079
the formula for calculating the sample entropy value is as follows:
Figure 636491DEST_PATH_IMAGE080
s216, differentiating the entropy value of the sample
Figure 988975DEST_PATH_IMAGE081
Initial single component of
Figure 133123DEST_PATH_IMAGE055
Adding to obtain new components
Figure 508740DEST_PATH_IMAGE082
Namely:
Figure 194937DEST_PATH_IMAGE083
in the formula,
Figure 588878DEST_PATH_IMAGE084
the component obtained by the octyl geometric modal decomposition is represented, and a represents the a-th component sequence.
In a specific embodiment, step S4 specifically includes:
s401, establishing an echo state network as follows:
Figure 702327DEST_PATH_IMAGE085
Figure 565241DEST_PATH_IMAGE086
Figure 789549DEST_PATH_IMAGE087
wherein,
Figure 303576DEST_PATH_IMAGE088
in order to input the dimension number, the dimension number is input,
Figure 994451DEST_PATH_IMAGE089
the number of internal neurons, l is the output dimension, and u (v), x (v) and y (v) are the input vector, the state vector and the output vector of the echo state network respectively;
s402, according to the input vector, the state vector and the output vector of the echo state network, training the echo state network by the following formula to obtain:
Figure 406978DEST_PATH_IMAGE090
Figure 887507DEST_PATH_IMAGE091
wherein f () is the internal neuron activation function Sigmoid, f out () As a function of the output layer(s),Wa connection weight matrix for internal states to internal states,
Figure 69089DEST_PATH_IMAGE092
for randomly generated input layers to a reserve pool
Figure 462025DEST_PATH_IMAGE093
×
Figure 286149DEST_PATH_IMAGE094
The order of the connection weight matrix is,
Figure 852259DEST_PATH_IMAGE095
for feeding back randomly generated output layers to reserve tanks
Figure 91611DEST_PATH_IMAGE093
A connection weight matrix of order x 1,
Figure 717764DEST_PATH_IMAGE096
is from reserve pool to output layer
Figure 26254DEST_PATH_IMAGE093
+
Figure 396056DEST_PATH_IMAGE094
+ l) order output weight matrix;
in which the echo-state network
Figure 489914DEST_PATH_IMAGE092
、WAnd
Figure 473919DEST_PATH_IMAGE095
all are randomly generated and are not changed in the learning process once generated, and only need to be adjusted in the training process of the reserve pool network
Figure 82755DEST_PATH_IMAGE096
The value of (2) is sufficient.
S403, optimizing the echo state network through a dimension competition slime algorithm, which specifically comprises the following steps:
1) Setting the total number U of individual slime organisms and the maximum iteration number of the dimension competition slime organism algorithm
Figure 662772DEST_PATH_IMAGE097
Proportional parameter z, decreasing parameter
Figure 673454DEST_PATH_IMAGE098
Random number
Figure 93939DEST_PATH_IMAGE099
Determining parameters of the updating method of the slime location
Figure 127755DEST_PATH_IMAGE100
Dimension D of individual slime bacteria, dimension competition probability Pv, and Gaussian variation probability
Figure 839359DEST_PATH_IMAGE101
；
2) Randomly generating a group of solutions as initial parameters to fit a dimension competition myxomycete algorithm to optimize an echo state network:
Figure 894427DEST_PATH_IMAGE102
wherein i =1,2.., U;
Figure 33284DEST_PATH_IMAGE103
in order to store the scale of the neurons in the pool,
Figure 554395DEST_PATH_IMAGE104
is the radius of the spectrum,
Figure 256641DEST_PATH_IMAGE105
in order to achieve the degree of sparseness,
Figure 976335DEST_PATH_IMAGE106
in order to input the dimensions of the cell,
Figure 223777DEST_PATH_IMAGE107
in order to displace the input unit, the displacement of the input unit,
Figure 294501DEST_PATH_IMAGE108
in order to output the dimensions of the cell,
Figure 534858DEST_PATH_IMAGE109
displacing the output unit;
3) Virtually exploring a target space through an initial parameter, and in t +1 iterations of the target space, updating the positions of the slime individuals in the following modes:
Figure 374638DEST_PATH_IMAGE110
Figure 527402DEST_PATH_IMAGE111
in the formula,
Figure 537952DEST_PATH_IMAGE112
and
Figure 129471DEST_PATH_IMAGE113
in order to search for the upper and lower boundaries of the range,
Figure 761440DEST_PATH_IMAGE114
is composed of
Figure 868461DEST_PATH_IMAGE115
Z is a ratio parameter for determining the ratio of randomly distributed slime bacteria individuals to slime bacteria,
Figure 648198DEST_PATH_IMAGE116
the position of the highest food odor concentration currently found, namely the optimal solution position;
Figure 981091DEST_PATH_IMAGE003
the current position of the slime mold;
Figure 795463DEST_PATH_IMAGE117
and
Figure 804876DEST_PATH_IMAGE116
the positions of two individuals randomly selected from the group are respectively;
Figure 9593DEST_PATH_IMAGE118
is the current iteration number;
Figure 208493DEST_PATH_IMAGE119
is a coefficient having a value of
Figure 64322DEST_PATH_IMAGE120
And gradually approaches 0 with the increase of the iteration number, wherein,
Figure 57686DEST_PATH_IMAGE121
，
Figure 484119DEST_PATH_IMAGE097
is the maximum iteration number;
Figure 221131DEST_PATH_IMAGE098
for a decreasing parameter from 1 to 0,
Figure 197046DEST_PATH_IMAGE099
is composed of
Figure 564574DEST_PATH_IMAGE115
A random number in between, and a random number,
Figure 806199DEST_PATH_IMAGE100
to determine the parameters of the updating method of the slime location,
Figure 259484DEST_PATH_IMAGE122
，
Figure 902955DEST_PATH_IMAGE123
is shown as
Figure 175805DEST_PATH_IMAGE059
The fitness value of each individual slime mold,
Figure 91677DEST_PATH_IMAGE124
expressing the optimal fitness value of the slime under the current iteration times;
the fitness value calculation formula is as follows:
Figure 170491DEST_PATH_IMAGE125
in the formula,
Figure 606152DEST_PATH_IMAGE126
、
Figure 377799DEST_PATH_IMAGE127
respectively predicting an actual value of the user side short-term power load and a predicted output value of the user side short-term power load; g is the number of training samples;
self-adaptive weight of slime bacteria
Figure 515388DEST_PATH_IMAGE128
The expression of (a) is:
Figure 69997DEST_PATH_IMAGE129
in the formula,
Figure 422481DEST_PATH_IMAGE099
is composed of
Figure 817559DEST_PATH_IMAGE115
A random number in between, and a random number,
Figure 255494DEST_PATH_IMAGE130
representing fitness values in ascending orderThe index of the individual position after the column,
Figure 613794DEST_PATH_IMAGE131
to represent
Figure 86364DEST_PATH_IMAGE132
The middle half of the top ranked population,
Figure 327377DEST_PATH_IMAGE133
represents the optimal fitness value obtained by the current iterative process,
Figure 439558DEST_PATH_IMAGE134
representing the worst fitness value obtained by the current iteration process;
4) For slime mold individual
Figure 601549DEST_PATH_IMAGE135
All dimensions are randomly paired without repeating pairwise pairs, and the pairs are D/2 pairs, and any pair of dimensions is paired, if rand<Pv, performing a dimension cross operator on the pair of dimensions according to the following formula;
Figure 663046DEST_PATH_IMAGE136
Figure 134348DEST_PATH_IMAGE137
；
Figure 546874DEST_PATH_IMAGE138
in the formula,
Figure 246977DEST_PATH_IMAGE139
is a slime individual
Figure 428560DEST_PATH_IMAGE135
To (1)
Figure 70763DEST_PATH_IMAGE140
And the first
Figure 908269DEST_PATH_IMAGE141
Dimension generation by dimension crossing;
Figure 474379DEST_PATH_IMAGE142
is [0,1 ]]A random number in between, and a random number,
Figure 965928DEST_PATH_IMAGE143
is the dimension cross probability;
it should be noted that, because the basic SMA algorithm is easy to have the problems of unstable optimization result, slow convergence speed, local optimization and the like when optimizing a high-dimensional complex function and a function of which the optimal solution is not at the origin, and the local optimization is often caused by the fact that one or more dimensions of the solution fall into the local optimization, a dimension competition operator is introduced into the slime mold algorithm.
The calculation amount of the dimension intersection operator can be controlled by setting different dimension intersection probability controls, so that the calculation speed of the algorithm is ensured.
5) Calculating progeny according to the formula
Figure 326502DEST_PATH_IMAGE144
With parent Myxomycetes individuals
Figure 651304DEST_PATH_IMAGE135
And updating the individual positions of the slime bacteria, and recording the current global optimal solution
Figure 208056DEST_PATH_IMAGE145
；
Figure 98652DEST_PATH_IMAGE146
6) If it is
Figure 833390DEST_PATH_IMAGE147
，
Figure 442226DEST_PATH_IMAGE148
Then entering a Gaussian mutation operator to carry out optimization on the optimal individuals
Figure 802669DEST_PATH_IMAGE145
Performing Gaussian mutation operation, further performing local search, and updating
Figure 485454DEST_PATH_IMAGE145
The position of (2):
Figure 453410DEST_PATH_IMAGE149
Figure 2072DEST_PATH_IMAGE150
in the formula,
Figure 713676DEST_PATH_IMAGE151
the particles are the optimal particles after Gaussian variation, N (0, 1) is a Gaussian distribution random quantity with the mean value of 0 and the variance of 1;
7) Judging the current iteration number
Figure 516547DEST_PATH_IMAGE118
Whether or not the maximum number of iterations has been reached
Figure 655404DEST_PATH_IMAGE097
If so, the iteration is finished, the optimal solution is output, otherwise,
Figure 428712DEST_PATH_IMAGE118
+1, return to step 2) and continue searching until the current iteration number
Figure 616111DEST_PATH_IMAGE118
To maximum number of iterations
Figure 601385DEST_PATH_IMAGE097
After iteration is stopped, outputting the current global optimal solution, and updating the echo by using the global optimal solutionAnd obtaining an initial power load prediction model according to the initial parameters of the state network.
The above is a detailed description of an embodiment of a short-term power load prediction method based on simmered geometry mode decomposition provided by the present invention, and the following is a detailed description of an embodiment of a short-term power load prediction system based on simmered geometry mode decomposition provided by the present invention.
For convenience of understanding, referring to fig. 2, the present invention provides a short-term power load prediction system based on simmer geometry decomposition, including:
the load acquisition module 100 is used for acquiring historical power loads of a user side and constructing a historical power load time sequence;
the decomposition module 200 is configured to decompose the historical power load time series by using a sine-shaped geometric mode to obtain a plurality of corresponding load components;
a dividing module 300 for dividing all load components into a training data set and a test data set;
the network construction module 400 is used for constructing an echo state network, and optimizing the echo state network through a dimension competition myxobacteria algorithm to obtain an initial power load prediction model;
the training module 500 is configured to input the load components in the training data set to the initial power load prediction model one by one for training, so as to obtain a power load prediction submodel corresponding to each load component;
the prediction module 600 is configured to input the test data set into the power load prediction submodel to perform prediction, so as to obtain a corresponding load prediction value;
and the superposition module 700 is configured to perform superposition processing on all the load predicted values to obtain a final user-side power load prediction result.
It can be clearly understood by those skilled in the art that, for convenience and simplicity of description, the specific working processes of the above-described systems, apparatuses and units may refer to the corresponding processes in the foregoing method embodiments, and are not described herein again.
In the embodiments provided in the present invention, it should be understood that the disclosed apparatus and method may be implemented in other ways. For example, the above-described apparatus embodiments are merely illustrative, and for example, the division of the units is only one type of logical functional division, and other divisions may be realized in practice, for example, multiple units or components may be combined or integrated into another system, or some features may be omitted, or not executed. In addition, the shown or discussed mutual coupling or direct coupling or communication connection may be an indirect coupling or communication connection through some interfaces, devices or units, and may be in an electrical, mechanical or other form.
The units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of the embodiment.
In addition, functional units in the embodiments of the present invention may be integrated into one processing unit, or each unit may exist alone physically, or two or more units are integrated into one unit. The integrated unit can be realized in a form of hardware, and can also be realized in a form of a software functional unit.
The above examples are only intended to illustrate the technical solution of the present invention, but not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.

Claims (4)

1. The short-term power load prediction method based on the octyl geometric modal decomposition is characterized by comprising the following steps of:
s1, obtaining historical power loads of a user side, and constructing a historical power load time sequence;
s2, decomposing the historical power load time sequence by adopting a sine geometric mode to obtain a plurality of corresponding load components;
s3, dividing all load components into a training data set and a testing data set;
s4, constructing an echo state network, and optimizing the echo state network through a dimension competition myxomycete algorithm to obtain an initial power load prediction model;
s5, inputting the load components in the training data set to the initial power load prediction model one by one for training to obtain power load prediction submodels corresponding to the load components;
s6, inputting the test data set into the power load forecasting submodel for forecasting to obtain a corresponding load forecasting value;
and S7, performing superposition processing on all the load predicted values to obtain a final user side power load prediction result.
2. The short-term power load forecasting method based on simpsoidal modal decomposition according to claim 1, wherein step S2 specifically includes:
s201, setting a historical power load time sequence as
Figure 427776DEST_PATH_IMAGE001
Wherein
Figure 436183DEST_PATH_IMAGE002
For the data length, the time sequence of the power load is reconstructed by adopting a time sequence delay topological equivalent method to obtain a track matrix
Figure 489589DEST_PATH_IMAGE003
Comprises the following steps:
Figure 516320DEST_PATH_IMAGE004
in the formula,
Figure 121614DEST_PATH_IMAGE005
in order to embed the dimension number, the number of the embedded dimension,
Figure 351738DEST_PATH_IMAGE006
in order to delay the time of the delay,
Figure 398716DEST_PATH_IMAGE007
；
s202, performing autocorrelation analysis on the track matrix X to obtain a covariance symmetric matrix
Figure 30686DEST_PATH_IMAGE008
Comprises the following steps:
Figure 869197DEST_PATH_IMAGE009
in the formula, T is a transposed symbol;
s203, a covariance symmetric matrix is matched
Figure 852197DEST_PATH_IMAGE008
Constructing a Hamilton matrix
Figure DEST_PATH_IMAGE010
Comprises the following steps:
Figure 434357DEST_PATH_IMAGE011
s204, aiming at the Hamilton matrix
Figure 655254DEST_PATH_IMAGE010
Squaring to obtain another Hamilton matrix
Figure 195825DEST_PATH_IMAGE012
Comprises the following steps:
Figure 400542DEST_PATH_IMAGE013
s205, constructing a sine orthogonal matrix by the following formula
Figure 789322DEST_PATH_IMAGE014
For Hamilton matrix
Figure 130305DEST_PATH_IMAGE012
And (3) carrying out octyl orthogonal transformation:
Figure 576199DEST_PATH_IMAGE015
in the formula,
Figure 268211DEST_PATH_IMAGE016
in the form of a real matrix,
Figure 192174DEST_PATH_IMAGE017
for upper triangular matrices, the matrix elements of an upper triangular matrix
Figure 715559DEST_PATH_IMAGE018
Wherein i and j are respectively the number of rows and columns of the matrix
Figure 817507DEST_PATH_IMAGE008
、
Figure 246083DEST_PATH_IMAGE017
And
Figure 724469DEST_PATH_IMAGE003
is expressed as:
Figure 557820DEST_PATH_IMAGE019
；
Figure 830670DEST_PATH_IMAGE020
in the form of a Householder matrix,
Figure 746542DEST_PATH_IMAGE021
wherein, in the process,
Figure 763040DEST_PATH_IMAGE022
，
Figure 447968DEST_PATH_IMAGE023
；
s206, calculating an upper triangular matrix by QR decomposition
Figure 954036DEST_PATH_IMAGE017
Characteristic value of (D) is noted
Figure 842357DEST_PATH_IMAGE024
From the upper triangular matrix
Figure 177392DEST_PATH_IMAGE017
Eigenvalue calculation matrix of
Figure 529876DEST_PATH_IMAGE008
The characteristic values of (A) are:
Figure 410108DEST_PATH_IMAGE025
for matrix
Figure 26204DEST_PATH_IMAGE008
The characteristic value of the vector is vectorized to obtain a corresponding characteristic vector which is recorded as
Figure 446821DEST_PATH_IMAGE026
；
S207, passing matrix
Figure 857074DEST_PATH_IMAGE008
Feature vector of
Figure 891895DEST_PATH_IMAGE026
And calculating the sum track matrix X to obtain a conversion coefficient matrix
Figure 817125DEST_PATH_IMAGE027
Comprises the following steps:
Figure 979116DEST_PATH_IMAGE028
using eigenvector matrices
Figure 493143DEST_PATH_IMAGE020
And a matrix of conversion coefficients
Figure 777494DEST_PATH_IMAGE027
Calculating to obtain a reconstruction matrix
Figure 862125DEST_PATH_IMAGE029
：
Figure 624544DEST_PATH_IMAGE030
Figure 258657DEST_PATH_IMAGE029
The initial single-component matrix is represented,
Figure 651592DEST_PATH_IMAGE031
，
Figure 551415DEST_PATH_IMAGE032
wherein the matrix is reconstructedZIs composed of
Figure 307406DEST_PATH_IMAGE033
Matrix of
Figure 281178DEST_PATH_IMAGE005
An initial single component matrix
Figure 907332DEST_PATH_IMAGE029
Is prepared from (A) and (B);
s208, for any initial single-component matrix
Figure 419084DEST_PATH_IMAGE029
Defining each element of the matrix as
Figure 523307DEST_PATH_IMAGE034
Wherein
Figure 866432DEST_PATH_IMAGE035
Let us order
Figure 335591DEST_PATH_IMAGE036
And an
Figure 944427DEST_PATH_IMAGE037
Wherein
Figure 570449DEST_PATH_IMAGE038
representing a single component matrix
Figure 581130DEST_PATH_IMAGE029
The minimum of the row and the column of (c),
Figure 486769DEST_PATH_IMAGE039
representing a single component matrix
Figure 772782DEST_PATH_IMAGE029
Of rows and columns, wherein when
Figure 218807DEST_PATH_IMAGE040
When it is taken
Figure 287257DEST_PATH_IMAGE041
When is coming into contact with
Figure 613065DEST_PATH_IMAGE042
Taking out
Figure 196493DEST_PATH_IMAGE043
Wherein
Figure 383892DEST_PATH_IMAGE044
represents the diagonal average transformation matrix calculation element values, the diagonal average transformation matrix is represented as follows:
Figure 556116DEST_PATH_IMAGE045
in the formula,
Figure 865874DEST_PATH_IMAGE046
representing the diagonal average transformation matrix calculation element values,
Figure 608702DEST_PATH_IMAGE047
and
Figure 114639DEST_PATH_IMAGE048
respectively representing the values of the calculated elements of the diagonal averaging transformation matrix
Figure 954419DEST_PATH_IMAGE047
And row and column
Figure 372762DEST_PATH_IMAGE048
The columns of the image data are,
Figure 855084DEST_PATH_IMAGE049
representing the kth element value of a diagonal averaging transformation matrix;
s209, converting the initial single component matrix by diagonal averaging
Figure 712181DEST_PATH_IMAGE029
Conversion to a one-dimensional time series of single-component signals, denoted as
Figure 344151DEST_PATH_IMAGE050
，
Figure 995712DEST_PATH_IMAGE051
Representing the data length, i is more than or equal to 1 and less than or equal to d, and sequentially performing matrix comparison on all initial single components
Figure 227979DEST_PATH_IMAGE052
Is obtained by carrying out diagonal averaging
Figure 560871DEST_PATH_IMAGE005
A one-dimensional time series, pair
Figure 109664DEST_PATH_IMAGE005
Superposing the one-dimensional time sequences to obtain a single-component total signal matrix as follows:
Figure 384657DEST_PATH_IMAGE053
wherein the single component total signal matrixYIs one
Figure 323794DEST_PATH_IMAGE054
A matrix;
s210, initial single component
Figure 709645DEST_PATH_IMAGE055
Form a group in sequence
Figure 378524DEST_PATH_IMAGE056
Vector of vitamin the vector is a function of the number of bits,
Figure 575150DEST_PATH_IMAGE057
Figure 77999DEST_PATH_IMAGE058
representing a single component from the origin
Figure 80590DEST_PATH_IMAGE055
To middle
Figure 541658DEST_PATH_IMAGE059
The points beginning to be continuous
Figure 892874DEST_PATH_IMAGE056
An
Figure 134500DEST_PATH_IMAGE060
A value of (d);
s211, defining vector
Figure 612885DEST_PATH_IMAGE058
And
Figure 177728DEST_PATH_IMAGE061
distance between them
Figure 778473DEST_PATH_IMAGE062
The element with the largest difference is the one with the largest difference between the two corresponding elements, namely:
Figure 913920DEST_PATH_IMAGE063
in the formula,
Figure 710843DEST_PATH_IMAGE061
representing a single component from the origin
Figure 146504DEST_PATH_IMAGE055
The j point in the figure begins continuously
Figure 652572DEST_PATH_IMAGE056
An
Figure 58670DEST_PATH_IMAGE060
The value of (a) is,
Figure 878858DEST_PATH_IMAGE064
；
Figure 231342DEST_PATH_IMAGE065
，
Figure 95262DEST_PATH_IMAGE066
；
s212, mixing
Figure 329934DEST_PATH_IMAGE062
Comparing with a preset threshold value, wherein the preset threshold value is 0.1 to 0.25
Figure 468660DEST_PATH_IMAGE067
Wherein
Figure 675651DEST_PATH_IMAGE067
determining standard deviation for single component signal
Figure 726783DEST_PATH_IMAGE062
The number of the distance is less than the preset threshold value, and the sum of the number and the distance is
Figure 841894DEST_PATH_IMAGE068
The ratio was calculated as:
Figure 3885DEST_PATH_IMAGE069
to find out their pairs
Figure 330961DEST_PATH_IMAGE056
Of vector of dimension average values are:
Figure 802263DEST_PATH_IMAGE070
s213. Dimension
Figure 886894DEST_PATH_IMAGE056
Plus 1, the sequence of the serial number arrangement forms a new group
Figure 101843DEST_PATH_IMAGE056
The + 1-dimensional vector is,
Figure 283426DEST_PATH_IMAGE071
Figure 410782DEST_PATH_IMAGE072
represents from the first
Figure 763135DEST_PATH_IMAGE059
The points beginning to be continuous
Figure 266928DEST_PATH_IMAGE056
+1 pieces
Figure 303017DEST_PATH_IMAGE060
A value of (d);
s214, defining a vector
Figure 119051DEST_PATH_IMAGE072
And
Figure 178274DEST_PATH_IMAGE073
distance between them
Figure 735026DEST_PATH_IMAGE074
The element with the largest difference is the one with the largest difference between the two corresponding elements, namely:
Figure 891201DEST_PATH_IMAGE075
in the formula,
Figure 625939DEST_PATH_IMAGE073
indicating continuation from the j-th point
Figure 421725DEST_PATH_IMAGE056
+1 pieces
Figure 1742DEST_PATH_IMAGE060
The value of (a) is set to (b),
Figure 464954DEST_PATH_IMAGE076
；
Figure 432910DEST_PATH_IMAGE065
，
Figure 466725DEST_PATH_IMAGE066
；
s215, mixing
Figure 356490DEST_PATH_IMAGE077
Comparing with a preset threshold value to determine
Figure 424941DEST_PATH_IMAGE077
The number of the distance is less than the preset threshold value, and the sum of the number and the distance is
Figure 563798DEST_PATH_IMAGE068
The ratio was calculated as:
Figure 68597DEST_PATH_IMAGE078
to find it
Figure 521575DEST_PATH_IMAGE056
The average of the +1 d vector vectors is worth:
Figure 693800DEST_PATH_IMAGE079
the formula for calculating the sample entropy value is as follows:
Figure 3558DEST_PATH_IMAGE080
s216, differentiating the entropy value of the sample
Figure 480807DEST_PATH_IMAGE081
Initial single component of
Figure 986744DEST_PATH_IMAGE055
Adding to obtain new components
Figure 764207DEST_PATH_IMAGE082
Namely:
Figure 244867DEST_PATH_IMAGE083
in the formula,
Figure 992767DEST_PATH_IMAGE084
representing a sine geometric modal decompositionThe resulting component, a, represents the a-th component sequence.
3. The short-term power load prediction method based on simutaneous geometric modal decomposition according to claim 1, wherein the step S4 specifically comprises:
s401, establishing an echo state network as follows:
Figure 521969DEST_PATH_IMAGE085
Figure 481835DEST_PATH_IMAGE086
Figure 258030DEST_PATH_IMAGE087
wherein,
Figure 37767DEST_PATH_IMAGE088
in order to input the dimension number, the dimension number is input,
Figure 619927DEST_PATH_IMAGE089
the number of internal neurons, l is the output dimension, and u (v), x (v) and y (v) are the input vector, the state vector and the output vector of the echo state network respectively;
s402, according to the input vector, the state vector and the output vector of the echo state network, training the echo state network by the following formula to obtain:
Figure 106403DEST_PATH_IMAGE090
Figure 381395DEST_PATH_IMAGE091
wherein f () is the internal neuron activation function Sigmoid, f out () As a function of the output layer(s),Wa connection weight matrix for internal states to internal states,
Figure 648428DEST_PATH_IMAGE092
for randomly generated input layers to a reserve pool
Figure 785012DEST_PATH_IMAGE093
×
Figure 643771DEST_PATH_IMAGE094
The order of the connection weight matrix is such that,
Figure 637135DEST_PATH_IMAGE095
for feeding back randomly generated output layers to reserve tanks
Figure 329147DEST_PATH_IMAGE093
A connection weight matrix of order x 1,
Figure 518689DEST_PATH_IMAGE096
is from reserve pool to output layer × (
Figure 979757DEST_PATH_IMAGE093
+
Figure 330973DEST_PATH_IMAGE094
+ l) order output weight matrix;
s403, optimizing the echo state network through a dimension competition myxomycete algorithm, which specifically comprises the following steps:
1) Setting the total number U of individual slime organisms and the maximum iteration number of the dimension competition slime organism algorithm
Figure 572599DEST_PATH_IMAGE097
Proportional parameter z, decreasing parameter
Figure 785405DEST_PATH_IMAGE098
Random number
Figure 615827DEST_PATH_IMAGE099
Determining parameters of the updating method of the slime location
Figure 216572DEST_PATH_IMAGE100
Dimension D of individual slime bacteria, dimension competition probability Pv, and Gaussian variation probability
Figure 617598DEST_PATH_IMAGE101
；
2) Randomly generating a group of solutions as initial parameters to fit a dimension competition myxomycete algorithm to optimize an echo state network:
Figure 151872DEST_PATH_IMAGE102
wherein i =1,2.., U;
Figure 321953DEST_PATH_IMAGE103
in order to store the scale of the neurons in the pool,
Figure 280551DEST_PATH_IMAGE104
in order to be the radius of the spectrum,
Figure 231189DEST_PATH_IMAGE105
in order to be sparse in degree,
Figure 51378DEST_PATH_IMAGE106
in order to input the dimensions of the cell,
Figure 590812DEST_PATH_IMAGE107
in order to input the displacement of the unit,
Figure 205464DEST_PATH_IMAGE108
in order to output the dimensions of the cell,
Figure 643399DEST_PATH_IMAGE109
displacing the output unit;
3) Virtually exploring a target space through an initial parameter, and in t +1 iterations of the target space, updating the positions of the slime individuals in the following modes:
Figure 782125DEST_PATH_IMAGE110
Figure 926799DEST_PATH_IMAGE111
in the formula,
Figure 40248DEST_PATH_IMAGE112
and
Figure 167078DEST_PATH_IMAGE113
in order to search the upper and lower boundaries of the range,
Figure 329069DEST_PATH_IMAGE114
is composed of
Figure 843096DEST_PATH_IMAGE115
Z is a ratio parameter for determining the ratio of randomly distributed slime bacteria individuals to slime bacteria,
Figure 861868DEST_PATH_IMAGE116
the position of the highest food odor concentration currently found, namely the optimal solution position;
Figure 212077DEST_PATH_IMAGE003
the current position of the slime mold;
Figure 427027DEST_PATH_IMAGE117
and
Figure 280713DEST_PATH_IMAGE116
the positions of two individuals randomly selected from the group are respectively;
Figure 735966DEST_PATH_IMAGE118
the current iteration number is;
Figure 822739DEST_PATH_IMAGE119
is a coefficient having a value of
Figure 326533DEST_PATH_IMAGE120
And gradually approaches 0 with the increase of the iteration number, wherein,
Figure 818082DEST_PATH_IMAGE121
，
Figure 444235DEST_PATH_IMAGE097
is the maximum iteration number;
Figure 503458DEST_PATH_IMAGE098
for a decreasing parameter from 1 to 0,
Figure 60210DEST_PATH_IMAGE099
is composed of
Figure 216385DEST_PATH_IMAGE115
A random number in between, and a random number,
Figure 951123DEST_PATH_IMAGE100
to determine the parameters of the updating method of the slime location,
Figure 746909DEST_PATH_IMAGE122
，
Figure 326926DEST_PATH_IMAGE123
representFirst, the
Figure 790137DEST_PATH_IMAGE059
The fitness value of each individual slime mold,
Figure 758093DEST_PATH_IMAGE124
representing the optimal fitness value of the slime mold under the current iteration times;
the fitness value calculation formula is as follows:
Figure 981789DEST_PATH_IMAGE125
in the formula,
Figure 693393DEST_PATH_IMAGE126
、
Figure 496264DEST_PATH_IMAGE127
respectively predicting an actual value of the user side short-term power load and a predicted output value of the user side short-term power load; g is the number of training samples;
slime self-adaptive weight
Figure 822072DEST_PATH_IMAGE128
The expression of (a) is:
Figure 405500DEST_PATH_IMAGE129
in the formula,
Figure 858478DEST_PATH_IMAGE099
is composed of
Figure 765123DEST_PATH_IMAGE115
A random number in between, and a random number,
Figure 746986DEST_PATH_IMAGE130
indicating the individual position indexes after the fitness values are arranged in ascending order,
Figure 817710DEST_PATH_IMAGE131
to represent
Figure 323646DEST_PATH_IMAGE132
The population of the first half of the middle rank,
Figure 101109DEST_PATH_IMAGE133
represents the optimal fitness value obtained by the current iteration process,
Figure 771650DEST_PATH_IMAGE134
representing the worst fitness value obtained by the current iteration process;
4) To slime bacteria individual
Figure 329670DEST_PATH_IMAGE135
All dimensions are randomly paired without repeating pairwise pairs, the total is D/2 pairs, and any pair of dimensions is paired, if rand<Pv, performing a dimension cross operator on the pair of dimensions according to the following formula;
Figure 858871DEST_PATH_IMAGE136
Figure 740109DEST_PATH_IMAGE137
；
Figure 594932DEST_PATH_IMAGE138
in the formula,
Figure 374669DEST_PATH_IMAGE139
is a slime mold individual
Figure 956829DEST_PATH_IMAGE135
To (1) a
Figure 443305DEST_PATH_IMAGE140
And the first
Figure 265768DEST_PATH_IMAGE141
Dimension generation by dimension crossing;
Figure 985331DEST_PATH_IMAGE142
is [0,1 ]]A random number in between, and a random number,
Figure 121914DEST_PATH_IMAGE143
is the dimension cross probability;
5) Calculating progeny according to the formula
Figure 968955DEST_PATH_IMAGE144
With parent Myxomycetes individuals
Figure 962319DEST_PATH_IMAGE135
And updating the individual positions of the slime bacteria, and recording the current global optimal solution
Figure 654331DEST_PATH_IMAGE145
；
Figure 578294DEST_PATH_IMAGE146
6) If it is
Figure 39362DEST_PATH_IMAGE147
，
Figure 469206DEST_PATH_IMAGE148
Then entering a Gaussian mutation operator to carry out optimization on the optimal individuals
Figure 897782DEST_PATH_IMAGE145
Performing Gaussian mutation operation, further performing local search, and updating
Figure 845010DEST_PATH_IMAGE145
The position of (c):
Figure 488481DEST_PATH_IMAGE149
Figure 276177DEST_PATH_IMAGE150
in the formula,
Figure 942782DEST_PATH_IMAGE151
the particles are the optimal particles after Gaussian variation, N (0, 1) is a Gaussian distribution random quantity with the mean value of 0 and the variance of 1;
7) Judging the current iteration number
Figure 211477DEST_PATH_IMAGE118
Whether or not the maximum number of iterations has been reached
Figure 709454DEST_PATH_IMAGE097
If so, the iteration is finished, the optimal solution is output, otherwise,
Figure 418784DEST_PATH_IMAGE118
+1, return to step 2) and continue searching until the current iteration number
Figure 494056DEST_PATH_IMAGE118
To a maximum number of iterations
Figure 376562DEST_PATH_IMAGE097
After iteration is stopped, outputting the current global optimal solution, and updating the initial parameters of the echo state network by using the global optimal solution to obtain the initial parametersA power load prediction model.
4. Short-term power load prediction system based on octyl geometric modal decomposition is characterized by comprising the following components:
the load acquisition module is used for acquiring historical power loads of a user side and constructing a historical power load time sequence;
the decomposition module is used for decomposing the historical power load time sequence by adopting a sine-shaped geometric mode to obtain a plurality of corresponding load components;
the dividing module is used for dividing all load components into a training data set and a testing data set;
the network construction module is used for constructing an echo state network, and optimizing the echo state network through a dimension competition myxobacteria algorithm to obtain an initial power load prediction model;
the training module is used for inputting the load components in the training data set to the initial power load prediction model one by one for training to obtain a power load prediction submodel corresponding to each load component;
the prediction module is used for inputting the test data set into the power load prediction submodel for prediction to obtain a corresponding load prediction value;
and the superposition module is used for carrying out superposition processing on all the load predicted values to obtain a final user side power load prediction result.
CN202211395000.7A 2022-11-09 2022-11-09 Short-term power load prediction method and system based on octyl geometric modal decomposition Pending CN115456312A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202211395000.7A CN115456312A (en) 2022-11-09 2022-11-09 Short-term power load prediction method and system based on octyl geometric modal decomposition

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202211395000.7A CN115456312A (en) 2022-11-09 2022-11-09 Short-term power load prediction method and system based on octyl geometric modal decomposition

Publications (1)

Publication Number Publication Date
CN115456312A true CN115456312A (en) 2022-12-09

Family

ID=84311095

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202211395000.7A Pending CN115456312A (en) 2022-11-09 2022-11-09 Short-term power load prediction method and system based on octyl geometric modal decomposition

Country Status (1)

Country Link
CN (1) CN115456312A (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115796231A (en) * 2023-01-28 2023-03-14 湖南赛能环测科技有限公司 Ultrashort-term wind speed prediction method based on temporal analysis
CN115860277A (en) * 2023-02-27 2023-03-28 西安骏硕通信技术有限公司 Data center energy consumption prediction method and system

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110705640A (en) * 2019-09-30 2020-01-17 温州大学 Method for constructing prediction model based on slime mold algorithm
US20200169085A1 (en) * 2017-06-28 2020-05-28 Silvio Becher Method for recognizing contingencies in a power supply network
CN113193556A (en) * 2021-05-12 2021-07-30 燕山大学 Short-term wind power prediction method based on probability prediction model

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20200169085A1 (en) * 2017-06-28 2020-05-28 Silvio Becher Method for recognizing contingencies in a power supply network
CN110705640A (en) * 2019-09-30 2020-01-17 温州大学 Method for constructing prediction model based on slime mold algorithm
CN113193556A (en) * 2021-05-12 2021-07-30 燕山大学 Short-term wind power prediction method based on probability prediction model

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
程正阳等: "辛几何模态分解方法及其分解能力研究", 《振动与冲击》 *

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115796231A (en) * 2023-01-28 2023-03-14 湖南赛能环测科技有限公司 Ultrashort-term wind speed prediction method based on temporal analysis
CN115796231B (en) * 2023-01-28 2023-12-08 湖南赛能环测科技有限公司 Temporal analysis ultra-short term wind speed prediction method
CN115860277A (en) * 2023-02-27 2023-03-28 西安骏硕通信技术有限公司 Data center energy consumption prediction method and system

Similar Documents

Publication Publication Date Title
CN112508275B (en) Power distribution network line load prediction method and equipment based on clustering and trend indexes
Tian et al. Multi-step short-term wind speed prediction based on integrated multi-model fusion
CN115456312A (en) Short-term power load prediction method and system based on octyl geometric modal decomposition
CN105354646B (en) Power load forecasting method for hybrid particle swarm optimization and extreme learning machine
CN110598854A (en) GRU model-based transformer area line loss rate prediction method
CN112733417B (en) Abnormal load data detection and correction method and system based on model optimization
CN113141008B (en) Data-driven power distribution network distributed new energy consumption capability assessment method
CN106022614A (en) Data mining method of neural network based on nearest neighbor clustering
CN113065704B (en) Super-parameter optimization and post-processing method of non-invasive load decomposition model
CN115115125B (en) Photovoltaic power interval probability prediction method based on deep learning fusion model
CN113780679B (en) Load prediction method and device based on ubiquitous power Internet of things
CN109492748A (en) A kind of Mid-long term load forecasting method for establishing model of the electric system based on convolutional neural networks
CN114219216B (en) Operation scheduling method, system, device and medium of power system
CN115964932A (en) Gas prediction method based on EMD-BilSTM-Attention mechanism transformer digital twin model
CN116169670A (en) Short-term non-resident load prediction method and system based on improved neural network
CN115409369A (en) Comprehensive energy system reliability evaluation method based on mechanism and data hybrid driving
CN115713144A (en) Short-term wind speed multi-step prediction method based on combined CGRU model
Wang et al. Causal carbon price interval prediction using lower upper bound estimation combined with asymmetric multi-objective evolutionary algorithm and long short-term memory
Ni et al. An improved attention-based bidirectional LSTM model for cyanobacterial bloom prediction
CN109408896B (en) Multi-element intelligent real-time monitoring method for anaerobic sewage treatment gas production
CN113762591B (en) Short-term electric quantity prediction method and system based on GRU and multi-core SVM countermeasure learning
CN114757107A (en) Gravity energy storage power distribution method based on load prediction model
Zhang et al. Optimal echo state network parameters based on behavioural spaces
Cao et al. Short-term forecasting and uncertainty analysis of photovoltaic power based on the FCM-WOA-BILSTM model
Wang et al. Sustainable investment forecasting of power grids based on the deep restricted Boltzmann machine optimized by the lion algorithm

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
RJ01 Rejection of invention patent application after publication
RJ01 Rejection of invention patent application after publication

Application publication date: 20221209