CN113988919A - Electricity purchasing structure adjusting method, system and medium - Google Patents

Electricity purchasing structure adjusting method, system and medium Download PDF

Info

Publication number
CN113988919A
CN113988919A CN202111242725.8A CN202111242725A CN113988919A CN 113988919 A CN113988919 A CN 113988919A CN 202111242725 A CN202111242725 A CN 202111242725A CN 113988919 A CN113988919 A CN 113988919A
Authority
CN
China
Prior art keywords
electricity purchasing
cost
electricity
similarity
factor
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
CN202111242725.8A
Other languages
Chinese (zh)
Inventor
陈浩
朱军飞
李京
王阳光
邓小亮
谢晓骞
李辉
张龙
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.)
State Grid Corp of China SGCC
State Grid Hunan Electric Power Co Ltd
Original Assignee
State Grid Corp of China SGCC
State Grid Hunan Electric Power 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 State Grid Corp of China SGCC, State Grid Hunan Electric Power Co Ltd filed Critical State Grid Corp of China SGCC
Priority to CN202111242725.8A priority Critical patent/CN113988919A/en
Publication of CN113988919A publication Critical patent/CN113988919A/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
    • G06Q30/00—Commerce
    • G06Q30/02—Marketing; Price estimation or determination; Fundraising
    • G06Q30/0201—Market modelling; Market analysis; Collecting market data
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00—Computer-aided design [CAD]
    • G06F30/20—Design optimisation, verification or simulation
    • G06F30/25—Design optimisation, verification or simulation using particle-based methods
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00—Computer-aided design [CAD]
    • G06F30/20—Design optimisation, verification or simulation
    • G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N20/00—Machine learning
    • G06N20/10—Machine learning using kernel methods, e.g. support vector machines [SVM]
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/004—Artificial life, i.e. computing arrangements simulating life
    • G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
    • 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
    • G06Q30/00—Commerce
    • G06Q30/02—Marketing; Price estimation or determination; Fundraising
    • G06Q30/0201—Market modelling; Market analysis; Collecting market data
    • G06Q30/0206—Price or cost determination based on market factors
    • 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—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
    • G06Q50/06—Electricity, gas or water supply
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2111/00—Details relating to CAD techniques
    • G06F2111/10—Numerical modelling

Abstract

The invention discloses a method, a system and a medium for adjusting a power purchase structure, wherein the method comprises the following steps: acquiring historical electricity purchasing data, wherein the historical electricity purchasing data comprises historical electricity purchasing cost and a historical electricity purchasing structure; selecting influence factors of the electricity purchasing cost according to the historical electricity purchasing data; processing the weighting coefficient of each influence factor by adopting a self-adaptive particle swarm optimization algorithm, inputting the processed data into a machine learning classification model which is constructed in advance and trained to predict the electricity purchasing cost, and obtaining the electricity purchasing prediction cost under the existing electricity purchasing structure; if the difference between the predicted cost of purchasing electricity and the set target cost exceeds the set value, the existing electricity purchasing structure is adjusted. The invention can better adjust the existing electricity purchasing structure to accurately make the electricity purchasing structure and reduce the electricity purchasing cost.

Description

Electricity purchasing structure adjusting method, system and medium
Technical Field
The invention relates to the electric power market technology, in particular to a method, a system and a medium for adjusting a power purchase structure.
Background
According to incomplete statistics, the electricity purchasing cost accounts for about 70% of all expenses of a power grid company. The electricity purchase cost has a high correlation with the electricity purchase structure. The accurate electricity purchasing cost prediction is beneficial to optimizing an electricity purchasing structure, consuming clean energy, optimizing energy configuration and the like of a power grid company, and is an important basis for controlling the cost and participating in market competition of the power grid company. But the electricity purchasing structure not only comprises the electricity generation of the traditional energy source but also comprises the electricity generation of the clean energy source. With the continuous expansion of the installed capacity of the clean energy, the market is occupied on a large scale by virtue of price advantage and policy inclination, the electricity purchasing cost of a power grid company is effectively reduced, however, the uncertainty of the clean energy increases the large-scale consumption difficulty, the consumption degree is obviously limited by weather and seasonal factors, and the electricity purchasing structure fluctuates month by month along with the consumption degree of the clean energy. In the actual production process, a power grid company establishes an electricity purchasing structure in advance, and allocates a monthly quota to each power plant or each party so as to provide the quota for each power plant or each party, but the electricity purchasing structure cannot be accurately established due to the uncertainty of clean energy, and the electricity purchasing cost is increased.
Disclosure of Invention
The technical problems to be solved by the invention are as follows: the invention aims to solve the problems that the power purchasing structure cannot be accurately established due to the uncertainty of clean energy and the power purchasing cost is increased.
In order to solve the technical problems, the invention adopts the technical scheme that:
a power purchase structure adjustment method comprises the following steps:
1) acquiring historical electricity purchasing data, wherein the historical electricity purchasing data comprises historical electricity purchasing cost and a historical electricity purchasing structure;
2) selecting influence factors of the electricity purchasing cost according to the historical electricity purchasing data;
3) processing the weighting coefficient of each influence factor by adopting a self-adaptive particle swarm optimization algorithm, inputting the processed data into a machine learning classification model which is constructed in advance and trained to predict the electricity purchasing cost, and obtaining the electricity purchasing prediction cost under the existing electricity purchasing structure;
4) if the difference between the predicted cost of purchasing electricity and the set target cost exceeds the set value, the existing electricity purchasing structure is adjusted.
Optionally, step 2) comprises:
2.1) obtaining initial influence factors from historical electricity purchasing data according to a preset rule;
2.2) calculating the comprehensive similarity between the electricity purchasing cost and the initial influence factors;
and 2.3) screening initial influence factors with the comprehensive similarity larger than a preset value as influence factors of the electricity purchasing cost.
Optionally, step 2.1) comprises:
2.1.1) obtaining initial influence factors from historical electricity purchasing data according to a preset rule;
2.1.2) forming an influence factor matrix by the data of each moment of each initial influence factor;
2.1.3) carrying out non-dimensionalization processing on the data in the influencing factor matrix;
2.1.4) calculating the weighting coefficient of each initial influence factor by adopting an entropy weight method;
2.1.5) the initial influencing factors whose weighting coefficients are below the threshold are removed.
Optionally, the method adopted for calculating the comprehensive similarity between the electricity purchase cost and the initial influence factor in step 2.2) is a gray correlation analysis method, and the steps include:
2.2.1) constructing an electricity purchasing cost matrix and an initial influence factor matrix according to the historical electricity purchasing cost and the initial influence factors;
2.2.2) constructing an electricity purchasing cost factor comprehensive matrix according to the electricity purchasing cost matrix and the initial influence factor matrix;
2.2.3) calculating the distance similarity and the shape similarity of the electricity purchasing cost and different initial influence factors;
2.2.4) obtaining the comprehensive similarity of the electricity purchasing cost and each initial influence factor according to the distance similarity and the shape similarity.
Optionally, when the distance similarity and the shape similarity between the electricity purchase cost and different initial influence factors are calculated in step 2.2.3), a calculation function expression of the distance similarity is as follows:
Figure BDA0003319791840000021
Δx1(i,j)＝|Xij-Xaj|，
Figure BDA0003319791840000022
in the above formula, γ1(i, j) is the distance similarity between the jth time value of the electricity purchase cost and the jth time value of the ith input element, and delta x1(i, j) is the shape difference between the jth time value of the electricity purchase cost and the jth time value of the ith input element, XijTo represent the value of the ith influencing factor at the jth moment, XajThe j time value of the electricity purchasing cost is X which is an integrated matrix of the electricity purchasing cost factors, Xa1,…,Xaj,…,XanRespectively as the electricity purchase cost matrix Xa1, …, j, …, n elements, X11～XmnRespectively representing the 1 st time value to the nth time value of the mth influencing factor of the 1 st influencing factor; i is 1,2, …, m, j is 1,2, …, n, m is the number of influencing factors, and n represents the nth time; the calculation function expression of the shape similarity is as follows:
Figure BDA0003319791840000023
in the above formula, γ2(i, j) is the shape similarity between the j-th time value of the electricity purchase cost and the j-th time value of the i-th input element, Δ x2(i, j) is the shape difference between the jth time value of the electricity purchase cost and the jth time value of the ith input element, XijTo represent the value of the ith influencing factor at the jth moment, XajThe j time value of the electricity purchasing cost is obtained.
Optionally, the calculation function expression of the comprehensive similarity between the electricity purchase cost and each initial influence factor obtained according to the distance similarity and the shape similarity in step 2.2.4) is as follows:
Figure BDA0003319791840000031
in the above formula, γiIndicates the integrated similarity, omega, of the ith influencing factoriA weighting factor, gamma, representing the ith influencing factor1(i, j) is the distance similarity between the j time value of the electricity purchase cost and the j time value of the i input element, gamma2(i, j) is the shape similarity between the j-th time value of the electricity purchase cost and the j-th time value of the i-th input element, and α and β are similarity coefficients and satisfy α + β being 1.
Optionally, when the weighting coefficient of each influence factor is processed by using the adaptive particle swarm optimization algorithm in step 3), the environment of the adaptive particle swarm optimization algorithm is to search in a D-dimensional space, a group is formed by N particles, and the expression of the particle update speed and the position is as follows:
Figure BDA0003319791840000032
Figure BDA0003319791840000033
in the above formula, the first and second carbon atoms are,
Figure BDA0003319791840000034
is the particle velocity at time t +1, ω is the weighting coefficient,
Figure BDA0003319791840000035
is the particle velocity at time t +1, c1And c2Is a learning factor, rand is [0,1 ]]A random number in between, and a random number,
Figure BDA0003319791840000036
for the optimal position that the particle itself has experienced,
Figure BDA0003319791840000037
is the position of the particle at time t,
Figure BDA0003319791840000038
for the optimal position that the particle population has experienced,
Figure BDA0003319791840000039
is the position of the particle at time t +1,
Figure BDA00033197918400000310
and
Figure BDA00033197918400000311
continuously updating in the iterative process, and finally outputting the optimal solution
Figure BDA00033197918400000312
The calculation function expression of the weighting coefficient ω is shown as follows:
Figure BDA00033197918400000313
in the above formula, ωmaxIs the maximum value of the weighting coefficient, ωminIs the minimum value of the weighting coefficient, favgAs the average value of the fitness of the particles, fmaxIs the maximum value of the fitness of the particle, fminIs the minimum value of the particle fitness, and f is the current particle fitness.
Optionally, the step 3) of inputting the processed data into a machine learning classification model which is constructed in advance and trained to predict the electricity purchasing cost includes: for sample set { (x)i,yi)|i＝1,2,…,N}，xiRepresenting a sample, xie.Rn, where Rn represents a set of real numbers, N represents the number of samples, yiE { -1,1} represents a sample label, and an optimal binary hyper-plane is established as shown in the following formula:
WΦ(x)+b＝0
in the above formula, W is a normal vector of the hyperplane, phi (x) is nonlinear transformation, and b is a constant term of the hyperplane;
through the optimal secondary classification hyperplane, the nonlinear feature vector is mapped into a plane space, the classification problem can be converted into an optimal secondary classification hyperplane optimization problem, and according to a risk minimization principle, the linear indivisible problem is converted into a linear constraint optimization problem:
Figure BDA0003319791840000041
in the above formula, L represents a Lagrangian function, N is the number of points, aiIs a Lagrangian factor, ajIs Lagrangian factor, yiThe category of the ith point is marked as 1 or-1, yjIs the category to which the jth point belongs, xiIs the feature vector of the ith point, xjThe feature vector of the j point; the machine learning classification model adopts an SVM prediction model, and a model decision function of the SVM prediction modelSelected as gaussian radial basis kernel function:
Figure BDA0003319791840000042
in the above equation, K (X, Y) is a gaussian radial basis function, X, Y are two vectors, | X-Y | is the distance between vectors X, Y, σ is a constant and σ ≠ 0.
In addition, the invention also provides a power purchase structure adjustment system, which comprises a microprocessor and a memory which are connected with each other, wherein the microprocessor is programmed or configured to execute the steps of the power purchase structure adjustment method, or the memory stores a computer program which is programmed or configured to execute the power purchase structure adjustment method.
Furthermore, the present invention also provides a computer-readable storage medium having stored therein a computer program programmed or configured to execute the electricity purchasing structure adjusting method.
Compared with the prior art, the invention has the following advantages: the method comprises the steps of firstly, acquiring historical electricity purchasing cost and historical electricity purchasing data of a historical electricity purchasing structure; then obtaining influence factors of electricity purchasing cost according to historical electricity purchasing data; then, obtaining the electricity purchasing cost prediction under the existing electricity purchasing structure according to the influence factors; and finally, adjusting the existing electricity purchasing structure according to the electricity purchasing cost prediction. According to the scheme, historical electricity purchasing data is provided, and influence factors influencing electricity purchasing cost are found; then, according to the influence factors, obtaining the electricity purchasing cost forecast under the existing electricity purchasing structure, and comparing the electricity purchasing cost forecast under the existing electricity purchasing structure with the expectation; and then better adjustment current electricity purchasing structure to accurate formulation electricity purchasing structure reduces the electricity purchasing cost.
Drawings
In order to more clearly illustrate the embodiments of the present application 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, it is obvious that the drawings in the following description are only some embodiments of the present application, and for those skilled in the art, other drawings can be obtained according to the drawings without creative efforts.
Fig. 1 is a flowchart of a method for adjusting a power purchase structure according to an embodiment of the present invention.
Fig. 2 is a specific flowchart of an electricity purchasing structure adjustment method according to an embodiment of the present invention.
Fig. 3 is a schematic structural diagram of an adjusting device for a power purchase structure according to an embodiment of the present invention.
Fig. 4 is a schematic structural diagram of an electricity purchasing structure adjustment system according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the following detailed description of the technical solutions of the present invention is provided with reference to the accompanying drawings and examples. It is to be understood that the embodiments described are only a few embodiments of the present application and not all embodiments. All other embodiments, which can be derived by a person skilled in the art from the examples given herein without making any creative effort, shall fall within the protection scope of the present application.
As shown in fig. 1, the method for adjusting the electricity purchasing structure of the present embodiment includes:
1) acquiring historical electricity purchasing data, wherein the historical electricity purchasing data comprises historical electricity purchasing cost and a historical electricity purchasing structure;
2) selecting influence factors of the electricity purchasing cost according to the historical electricity purchasing data;
3) processing the weighting coefficient of each influence factor by adopting a self-adaptive particle swarm optimization algorithm, inputting the processed data into a machine learning classification model which is constructed in advance and trained to predict the electricity purchasing cost, and obtaining the electricity purchasing prediction cost under the existing electricity purchasing structure;
4) if the difference between the predicted cost of purchasing electricity and the set target cost exceeds the set value, the existing electricity purchasing structure is adjusted.
The method of the embodiment finds out the influence factors influencing the electricity purchasing cost based on the provided historical electricity purchasing data; then, according to the influence factors, obtaining the electricity purchasing cost forecast under the existing electricity purchasing structure, and comparing the electricity purchasing cost forecast under the existing electricity purchasing structure with the expectation; and then better adjustment current electricity purchasing structure to accurate formulation electricity purchasing structure reduces the electricity purchasing cost.
As shown in fig. 1, step 1) in this embodiment is a step of data preparation and processing. And acquiring the power generation ratio of various power supplies changing along with time and a power purchase contract signed by a power grid company and a power plant in a corresponding period of time through a data interface between the power supply and a contract management system of a related system and a related organization, and acquiring historical power purchase cost and historical power purchase structure data of the power grid company.
In this embodiment, step 2) uses the historical electricity purchasing cost and the electricity purchasing structure as influencing factors according to the collected historical data, wherein the electricity purchasing structure is determined by the power type in the province and the external electricity purchasing, and therefore the electricity purchasing structure influencing factors can be subdivided according to the specific installation situation of each province. Finally, based on the actual conditions of each province, the factors influencing the electricity purchasing cost are obtained, and the entropy weight method is adopted to carry out weighting processing on each influencing factor so as to clarify the contribution degree of different influencing factors.
In this embodiment, step 2) includes:
2.1) obtaining initial influence factors from historical electricity purchasing data according to a preset rule;
2.2) calculating the comprehensive similarity between the electricity purchasing cost and the initial influence factors;
and 2.3) screening initial influence factors with the comprehensive similarity larger than a preset value as influence factors of the electricity purchasing cost.
In this embodiment, step 2.1) includes:
2.1.1) obtaining initial influence factors from historical electricity purchasing data according to a preset rule; it should be noted that the preset rule here refers to an extraction rule matched with the data format of the historical electricity purchasing data, and is used for extracting corresponding influence factor information;
2.1.2) forming an influence factor matrix by the data of each moment of each initial influence factor;
2.1.3) carrying out non-dimensionalization processing on the data in the influencing factor matrix;
2.1.4) calculating the weighting coefficient of each initial influence factor by adopting an entropy weight method;
2.1.5) the initial influencing factors whose weighting coefficients are below the threshold are removed.
Because the influence factors are various and the respective dimensions are different, in order to avoid reduction of load prediction precision caused by dimension difference, in step 2.1.3), when carrying out non-dimensionalization processing on the data in the influence factor matrix, carrying out non-dimensionalization processing on the acquired original data by adopting a minimum value and maximum value method, and simultaneously carrying out data processing by adopting an average interpolation method. The functional expression of the minimum and maximum methods dimensionless process is:
Figure BDA0003319791840000061
in the above formula, xiThe numerical values are subjected to non-dimensionalization;
Figure BDA0003319791840000062
the value of the original numerical value is obtained,
Figure BDA0003319791840000063
is the jth influence factor matrix minimum;
Figure BDA0003319791840000064
is the jth largest value of the influencing factor matrix.
Step 2.1.4) when the weighting coefficient of each initial influence factor is calculated by adopting an entropy weight method, the function expression is as follows:
Figure BDA0003319791840000065
β＝1/lnn，
in the above formula, ωiIs a weighting coefficient of the ith influencing factor, HiThe degree of dispersion of indexes can be effectively measured for the information entropy in the information theory; m is the number of influencing factors; beta is the information entropy proportionality coefficient, fikIs the probability of occurrence of the ith influencing factor, xikThe numerical value of the ith influencing factor in the k dimension and n at the nth moment need to be explained, wherein the historical electricity purchasing number is used for constructing the matrixThe data of each month is used as data of a moment to construct, and the moment represented by n is actually the month of the data. By calculating the weighting coefficient, the initial influence factors with low weight are deleted in advance, and the subsequent calculation process can be lightened.
In this embodiment, the method adopted for calculating the comprehensive similarity between the electricity purchase cost and the initial influence factor in step 2.2) is a gray correlation analysis method, and includes the steps of:
2.2.1) constructing an electricity purchasing cost matrix and an initial influence factor matrix according to the historical electricity purchasing cost and the initial influence factors;
2.2.2) constructing an electricity purchasing cost factor comprehensive matrix according to the electricity purchasing cost matrix and the initial influence factor matrix;
2.2.3) calculating the distance similarity and the shape similarity of the electricity purchasing cost and different initial influence factors;
2.2.4) obtaining the comprehensive similarity of the electricity purchasing cost and each initial influence factor according to the distance similarity and the shape similarity.
The traditional grey correlation analysis mainly measures the shape similarity between sequences, does not consider the distance difference of different sequences, and therefore may have obvious difference from the fact. In order to ensure that the selected influence factors have a strong correlation with the electricity purchase cost in practice, the influence factors of the input model are selected by the improved gray correlation analysis shown in the foregoing steps 2.2.1) -2.2.4) in this embodiment by using a comprehensive similarity calculation method considering the combination of the shape and the distance.
In this embodiment, when the distance similarity and the shape similarity between the electricity purchase cost and different initial influence factors are calculated in step 2.2.3), a calculation function expression of the distance similarity is as follows:
Figure BDA0003319791840000071
Δx1(i,j)＝|Xij-Xaj|，
Figure BDA0003319791840000072
in the above formula, γ1(i, j) is the distance similarity between the jth time value of the electricity purchase cost and the jth time value of the ith input element, and delta x1(i, j) is the shape difference between the jth time value of the electricity purchase cost and the jth time value of the ith input element, XijTo represent the value of the ith influencing factor at the jth moment, XajThe j time value of the electricity purchasing cost is X which is an integrated matrix of the electricity purchasing cost factors, Xa1,…,Xaj,…,XanRespectively as the electricity purchase cost matrix Xa1, …, j, …, n elements, X11～XmnRespectively representing the 1 st time value to the nth time value of the mth influencing factor of the 1 st influencing factor; i is 1,2, …, m, j is 1,2, …, n, m is the number of influencing factors, and n represents the nth time; the calculation function expression of the shape similarity is as follows:
Figure BDA0003319791840000073
in the above formula, γ2(i, j) is the shape similarity between the j-th time value of the electricity purchase cost and the j-th time value of the i-th input element, Δ x2(i, j) is the shape difference between the jth time value of the electricity purchase cost and the jth time value of the ith input element, XijTo represent the value of the ith influencing factor at the jth moment, XajThe j time value of the electricity purchasing cost is obtained.
In this embodiment, the calculation function expression of the comprehensive similarity between the electricity purchase cost and each initial influence factor obtained in step 2.2.4) according to the distance similarity and the shape similarity is as follows:
Figure BDA0003319791840000081
in the above formula, γiIndicates the integrated similarity, omega, of the ith influencing factoriA weighting factor, gamma, representing the ith influencing factor1(i, j) is the distance phase between the jth time value of the electricity purchase cost and the jth time value of the ith input elementSimilarity, γ2(i, j) is the shape similarity between the j-th time value of the electricity purchase cost and the j-th time value of the i-th input element, and α and β are similarity coefficients and satisfy α + β being 1. The values of α and β depend on the importance of distance similarity and shape similarity. In this embodiment, the importance of the distance similarity and the importance of the shape similarity are not in order, and therefore α ═ β ═ 0.5. And setting a comprehensive similarity threshold, and selecting the influence factors exceeding the threshold as the influence factors of the prediction model.
In the embodiment, step 3) is to form power grid company power purchase cost prediction based on an APSO-SVM model, the APSO-SVM model comprises an adaptive particle swarm optimization algorithm and an SVM model, the power grid company power purchase cost prediction changing along with time change is obtained by using historical data of influence factors of the prediction model selected in step 2) as input data based on the APSO-SVM model, the method is to process the weighting coefficient of each influence factor by adopting the adaptive particle swarm optimization algorithm, and the processed data is input into the SVM model which is constructed in advance and trained to predict the power purchase cost, so that the power purchase prediction cost under the existing power purchase structure is obtained. The SVM model may also adopt other machine learning classification models as needed, and similarly, the electricity purchase cost prediction of the processed data may also be implemented.
In this embodiment, the fitness of each individual in the population to the environment is analyzed through a weight-dynamic Adaptive Particle Swarm Optimization (APSO) algorithm, a position with a better area in the individual is found, and the individual in the population is moved to the best position in the area by utilizing mutual cooperation and information sharing. Namely, the selected historical data of the influencing factors are optimized, and the obtained optimized data is used as the input data of the prediction model, so that the output result of the prediction model is more accurate. Specifically, when the adaptive particle swarm optimization algorithm is adopted to process the weighting coefficient of each influence factor in step 3), the environment of the adaptive particle swarm optimization algorithm is to search in a D-dimensional space, a group is formed by N particles, and the expression of the particle update speed and the position is as follows:
Figure BDA0003319791840000082
Figure BDA0003319791840000083
in the above formula, the first and second carbon atoms are,
Figure BDA0003319791840000084
is the particle velocity at time t +1, ω is the weighting coefficient,
Figure BDA0003319791840000085
is the particle velocity at time t +1, c1And c2Is a learning factor, rand is [0,1 ]]A random number in between, and a random number,
Figure BDA0003319791840000086
for the optimal position that the particle itself has experienced,
Figure BDA0003319791840000087
is the position of the particle at time t,
Figure BDA0003319791840000088
for the optimal position that the particle population has experienced,
Figure BDA0003319791840000089
is the position of the particle at time t +1,
Figure BDA00033197918400000810
and
Figure BDA00033197918400000811
continuously updating in the iterative process, and finally outputting the optimal solution
Figure BDA00033197918400000812
The calculation function expression of the weighting coefficient ω is shown as follows:
Figure BDA0003319791840000091
in the above formula, ωmaxIs the maximum value of the weighting coefficient, ωminIs the minimum value of the weighting coefficient, favgAs the average value of the fitness of the particles, fmaxIs the maximum value of the fitness of the particle, fminIs the minimum value of the particle fitness, and f is the current particle fitness.
In the embodiment, the prediction of the power purchase cost of the power grid company is output by improving the SVM prediction model of the particle swarm optimization algorithm. The model firstly maps the nonlinear feature vectors into a plane space, and effectively finishes two classifications by utilizing a hyperplane mode. Specifically, the step 3) of inputting the processed data into a machine learning classification model which is constructed in advance and trained to predict the electricity purchasing cost comprises the following steps: for sample set { (x)i,yi)|i＝1,2,…,N}，xiRepresenting a sample, xie.Rn, where Rn represents a set of real numbers, N represents the number of samples, yiE { -1,1} represents a sample label, and an optimal binary hyper-plane is established as shown in the following formula:
WΦ(x)+b＝0
in the above formula, W is a normal vector of the hyperplane, phi (x) is nonlinear transformation, and b is a constant term of the hyperplane; at this point, the classification problem can be converted to an optimal two-classification hyperplane optimization problem. Through the optimal two-classification hyperplane, the nonlinear feature vector is mapped into a plane space, the classification problem (linear non-separable problem) can be converted into an optimal two-classification hyperplane optimization problem, and the linear non-separable problem is converted into a linear constraint optimization problem according to a risk minimization principle:
Figure BDA0003319791840000092
in the above formula, L represents a Lagrangian function, N is the number of points, aiIs a Lagrangian factor, ajIs Lagrangian factor, yiThe category of the ith point is marked as 1 or-1, yjIs the category to which the jth point belongs, xiIs the ithFeature vector of points, xjThe feature vector of the j point; the corresponding training samples are called support vector machines and denoted as ai. As described above, in this embodiment, the machine learning classification model adopts an SVM prediction model, and the model decision function of the SVM prediction model is selected as a gaussian radial basis kernel function:
Figure BDA0003319791840000093
in the above equation, K (X, Y) is a gaussian radial basis function, X, Y are two vectors, | X-Y | is the distance between vectors X, Y, σ is a constant and σ ≠ 0.
The electricity purchasing cost of each fixed time period such as each month can be predicted through the SVM model, and the obtained electricity purchasing cost prediction of each month is compared with the electricity purchasing cost budget per se; if the electricity purchasing cost prediction is larger than the self electricity purchasing cost budget; or even if the electricity purchasing cost forecast is smaller than the electricity purchasing cost budget per se, the electricity purchasing cost is further reduced; the adjustment can be performed according to preset rules. For example, if the capacity of the clean energy installation can be increased, the electricity purchasing proportion of the clean energy electric energy can be increased in the corresponding month. If the capacity of clean energy in the province or the region can not be increased in the month, but the price of the externally purchased electricity is cheaper than that of the traditional energy power generation, and the occupation ratio of the externally purchased electricity can be increased. Or establishing a model, and inputting the desired electricity purchasing cost into the model to obtain the optimal electricity purchasing structure under the electricity purchasing cost.
In the embodiments of the present invention, the power grid power purchase structure adjustment is described, but in actual cases, the power purchase structure of the power grid corresponds to the power generation amount of each power plant and the outsourcing power problem. Therefore, the embodiment of the invention also discloses a control method for controlling the monthly power generation amount of each power station, which obtains the monthly power generation amount of each power station according to the predicted power purchase cost and controls each power station to generate power according to the corresponding power generation amount every month. The most important scheme is the prediction of electricity purchasing cost, which is consistent with the scheme provided by the embodiment of the invention and is not described herein again.
In order to more clearly illustrate the cost prediction scheme in the embodiment of the invention, a specific case is provided; in order to verify the effectiveness and the accuracy of the power purchase cost prediction method considering the change of the power purchase structure, the power purchase structure and the actual data of the power purchase cost of a certain power-saving network company are adopted for analysis.
The first step is as follows: through a data interface between the system and a D5000 power system and a contract management system of a related organization, the influence factors of the electricity purchasing structures and historical data of electricity purchasing cost from 2011 to 2013 of a certain province are collected as samples. The sampling time interval of the sample is 1 month, 12 data points are acquired every year, and 36 data points (corresponding to the time represented by n in the embodiment of the invention) are counted, wherein the first 30 data points are selected as training samples of the SVM model, and the last 6 data points are selected as testing samples of the SVM model. Since there is more data, it is not listed here.
The second step is that: the system determines the power type of the province according to the collected data, selects the obtained power type and the outsourcing electricity and electricity purchasing cost as initial influence factors, and performs weighting processing on the initial influence factors by adopting an entropy weight method. Meanwhile, a comprehensive similarity value of different influence factors and the predicted electricity purchasing cost is calculated by using an improved grey correlation analysis method.
In this case, the initial influencing factors selected by the system include: thermal power, hydroelectric power, nuclear power, gas power, other power generation, outsourcing power and historical power purchase costs. Then, corresponding comprehensive similarity is obtained through calculation, and the comprehensive similarity is respectively as follows: 0.6519, 0.5743, 0.7548, 0.9037, 0.5307, 0.8132, and 0.8983. The gas electricity, outsourcing electricity and historical electricity purchasing cost with the comprehensive similarity exceeding a set threshold are selected as influence factors of the electricity purchasing cost prediction model, and relevant historical data of the gas electricity, outsourcing electricity and historical electricity purchasing cost are used as input data of the prediction model.
The third step: and inputting the data, and finally forming the power purchase cost prediction of the power grid company through an APSO-SVM power purchase cost prediction model.
Under the double-carbon background and the electric power market conditions in China, clean energy gradually participates in the market, and the proportion is continuously increased. However, the clean energy power generation has high uncertainty, which results in large changes of the power purchasing structure, and the power purchasing cost of the power grid company is easy to change suddenly. Therefore, in the province with higher proportion of clean energy, the electricity purchasing structure of the clean energy obviously changes along with time, and certain influence is caused on the electricity purchasing cost of a power grid company. In order to enable a power grid company to better consume clean energy, optimize energy configuration and electricity purchasing structure and control cost, accurate prediction of electricity purchasing cost of the power grid company is needed. Then, the existing electricity purchasing structure is adjusted in time, and the electricity purchasing cost is reduced.
The fourth step: and adjusting the existing electricity purchasing structure according to the electricity purchasing cost prediction.
In one embodiment, as shown in fig. 3, an embodiment of the present invention provides an electricity purchasing structure adjusting apparatus, including:
a historical electricity purchasing data acquiring module 31, configured to acquire historical electricity purchasing data, where the historical electricity purchasing data includes historical electricity purchasing costs and a historical electricity purchasing structure;
the influence factor acquisition module 32 is used for acquiring influence factors of the electricity purchasing cost according to the historical electricity purchasing data; specifically, the influence factor obtaining module 32 obtains an initial influence factor from historical electricity purchasing data according to a preset rule; calculating the comprehensive similarity between the electricity purchasing cost and the initial influence factors; and screening the initial influence factors with the comprehensive similarity larger than a preset value as influence factors.
The electricity purchasing cost prediction module 33 is used for obtaining electricity purchasing cost prediction under the existing electricity purchasing structure according to the influence factors; specifically, the electricity purchasing cost prediction module 33 processes the weighting coefficient of each influence factor by using a self-adaptive particle swarm optimization algorithm; and inputting the processed data into a pre-constructed SVM model to obtain the electricity purchasing cost prediction under the existing electricity purchasing structure.
And the electricity purchasing structure adjusting module 34 is used for adjusting the existing electricity purchasing structure according to the electricity purchasing cost prediction.
According to the adjusting device for the electricity purchasing structure provided by the embodiment of the invention, the historical electricity purchasing data acquiring module acquires historical electricity purchasing data; the influence factor acquisition module acquires influence factors of the electricity purchasing cost according to historical electricity purchasing data; the electricity purchasing cost prediction module obtains electricity purchasing cost prediction under the existing electricity purchasing structure according to the influence factors; the existing electricity purchasing structure adjusting module adjusts the existing electricity purchasing structure according to the electricity purchasing cost prediction. The device acquires required historical data and outputs the data processed by the non-quantitative tempering and average interpolation method. Analyzing factors influencing the electricity purchasing structure according to input data, and performing weighting processing on the factors by using an entropy weight method; then selecting influence factors in the prediction model by improving grey correlation analysis, namely determining the factor with the largest influence by calculating distance similarity, shape similarity and comprehensive similarity; and finally outputting the historical data of the selected influencing factors. And constructing an effective and accurate electricity purchasing cost prediction model and solving to form the electricity purchasing cost prediction of the power grid company based on the electricity purchasing structure changing along with time. The input data are optimized by a self-adaptive particle swarm algorithm and then input into an SVM prediction model, namely an APSO-SVM prediction model; and finally, outputting the power purchase cost prediction of the power grid company.
In one embodiment, as shown in fig. 4, the present invention further provides an electricity purchasing structure adjusting system, including:
the processor(s) 41 are (are) arranged,
a memory 42 for storing instructions executable by the processor 41;
wherein the processor 41 is programmed or configured to perform the steps of the aforementioned power purchase configuration adjustment method.
In addition, the present embodiment further provides an electricity purchasing structure adjusting system, which includes a microprocessor and a memory, which are connected to each other, wherein the microprocessor is programmed or configured to execute the steps of the electricity purchasing structure adjusting method, or the memory stores a computer program programmed or configured to execute the electricity purchasing structure adjusting method.
In addition, the present embodiment also provides a computer-readable storage medium, in which a computer program programmed or configured to execute the foregoing electricity purchasing structure adjusting method is stored.
It is understood that the same or similar parts in the above embodiments may be mutually referred to, and the same or similar parts in other embodiments may be referred to for the content which is not described in detail in some embodiments.
It should be noted that, in the description of the present application, the terms "first", "second", etc. are used for descriptive purposes only and are not to be construed as indicating or implying relative importance. Further, in the description of the present application, the meaning of "a plurality" means at least two unless otherwise specified.
Any process or method descriptions in flow charts or otherwise described herein may be understood as representing modules, segments, or portions of code which include one or more executable instructions for implementing specific logical functions or steps of the process, and the scope of the preferred embodiments of the present application includes other implementations in which functions may be executed out of order from that shown or discussed, including substantially concurrently or in reverse order, depending on the functionality involved, as would be understood by those reasonably skilled in the art of the present application.
It should be understood that portions of the present application may be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, the various steps or methods may be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, any one or combination of the following techniques, which are known in the art, may be used: a discrete logic circuit having a logic gate circuit for implementing a logic function on a data signal, an application specific integrated circuit having an appropriate combinational logic gate circuit, a Programmable Gate Array (PGA), a Field Programmable Gate Array (FPGA), or the like.
It will be understood by those skilled in the art that all or part of the steps carried by the method for implementing the above embodiments may be implemented by hardware associated with program instructions, and the program may be stored in a computer readable storage medium, and when executed, the program includes one or a combination of the steps of the method embodiments.
In addition, functional units in the embodiments of the present application may be integrated into one processing module, or each unit may exist alone physically, or two or more units are integrated into one module. The integrated module can be realized in a hardware mode, and can also be realized in a software functional module mode. The integrated module, if implemented in the form of a software functional module and sold or used as a stand-alone product, may also be stored in a computer readable storage medium.
The storage medium mentioned above may be a read-only memory, a magnetic or optical disk, etc.
In the description herein, reference to the description of the term "one embodiment," "some embodiments," "an example," "a specific example," or "some examples," etc., means that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the application. In this specification, the schematic representations of the terms used above do not necessarily refer to the same embodiment or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.
Although embodiments of the present application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present application, and that variations, modifications, substitutions and alterations may be made to the above embodiments by those of ordinary skill in the art within the scope of the present application.

Claims (10)

1. A power purchase structure adjustment method is characterized by comprising the following steps:
1) acquiring historical electricity purchasing data, wherein the historical electricity purchasing data comprises historical electricity purchasing cost and a historical electricity purchasing structure;
2) selecting influence factors of the electricity purchasing cost according to the historical electricity purchasing data;
3) processing the weighting coefficient of each influence factor by adopting a self-adaptive particle swarm optimization algorithm, inputting the processed data into a machine learning classification model which is constructed in advance and trained to predict the electricity purchasing cost, and obtaining the electricity purchasing prediction cost under the existing electricity purchasing structure;
4) if the difference between the predicted cost of purchasing electricity and the set target cost exceeds the set value, the existing electricity purchasing structure is adjusted.
2. The power purchase structure adjustment method according to claim 1, wherein the step 2) includes:
2.1) obtaining initial influence factors from historical electricity purchasing data according to a preset rule;
2.2) calculating the comprehensive similarity between the electricity purchasing cost and the initial influence factors;
and 2.3) screening initial influence factors with the comprehensive similarity larger than a preset value as influence factors of the electricity purchasing cost.
3. The power purchase structure adjustment method according to claim 2, wherein step 2.1) comprises:
2.1.1) obtaining initial influence factors from historical electricity purchasing data according to a preset rule;
2.1.2) forming an influence factor matrix by the data of each moment of each initial influence factor;
2.1.3) carrying out non-dimensionalization processing on the data in the influencing factor matrix;
2.1.4) calculating the weighting coefficient of each initial influence factor by adopting an entropy weight method;
2.1.5) the initial influencing factors whose weighting coefficients are below the threshold are removed.
4. The method for adjusting the electricity purchasing structure according to claim 2, wherein the method for calculating the comprehensive similarity between the electricity purchasing cost and the initial influence factors in the step 2.2) is a grey correlation analysis method, and the method comprises the following steps:
2.2.1) constructing an electricity purchasing cost matrix and an initial influence factor matrix according to the historical electricity purchasing cost and the initial influence factors;
2.2.2) constructing an electricity purchasing cost factor comprehensive matrix according to the electricity purchasing cost matrix and the initial influence factor matrix;
2.2.3) calculating the distance similarity and the shape similarity of the electricity purchasing cost and different initial influence factors;
2.2.4) obtaining the comprehensive similarity of the electricity purchasing cost and each initial influence factor according to the distance similarity and the shape similarity.
5. The power purchase structure adjustment method according to claim 4, wherein, when calculating the distance similarity and the shape similarity between the power purchase cost and different initial influencing factors in step 2.2.3), the calculation function expression of the distance similarity is as follows:
Figure FDA0003319791830000011
Δx1(i，j)＝|Xij-Xaj|，
Figure FDA0003319791830000012
in the above formula, γ1(i, j) is the distance similarity between the jth time value of the electricity purchase cost and the jth time value of the ith input element, and delta x1(i, j) is the shape difference between the jth time value of the electricity purchase cost and the jth time value of the ith input element, XijTo represent the value of the ith influencing factor at the jth moment, XajThe j time value of the electricity purchasing cost is X which is an integrated matrix of the electricity purchasing cost factors, Xa1,…,Xaj,…,XanRespectively as the electricity purchase cost matrix Xa1, …, j, …, n elements, X11～XmnRespectively representing the 1 st time value to the nth time value of the mth influencing factor of the 1 st influencing factor; i is 1,2, …, m, j is 1,2, …, n, m is the number of influencing factors, and n represents the nth time; the calculation function expression of the shape similarity is as follows:
Figure FDA0003319791830000021
in the above formula, γ2(i, j) is the shape similarity between the j-th time value of the electricity purchase cost and the j-th time value of the i-th input element, Δ x2(i, j) is the j time value of the electricity purchase cost and the j time of the ith input elementShape difference of scale value, XijTo represent the value of the ith influencing factor at the jth moment, XajThe j time value of the electricity purchasing cost is obtained.
6. The method for adjusting the electricity purchasing structure according to claim 5, wherein the calculation function expression of the comprehensive similarity between the electricity purchasing cost and each initial influence factor obtained according to the distance similarity and the shape similarity in step 2.2.4) is as follows:
Figure FDA0003319791830000022
in the above formula, γiIndicates the integrated similarity, omega, of the ith influencing factoriA weighting factor, gamma, representing the ith influencing factor1(i, j) is the distance similarity between the j time value of the electricity purchase cost and the j time value of the i input element, gamma2(i, j) is the shape similarity between the j-th time value of the electricity purchase cost and the j-th time value of the i-th input element, and α and β are similarity coefficients and satisfy α + β being 1.
7. The power purchase structure adjustment method according to claim 6, wherein when the adaptive particle swarm optimization algorithm is used to process the weighting coefficient of each influence factor in step 3), the environment of the adaptive particle swarm optimization algorithm is to search in a D-dimensional space, a group is formed by N particles, and the expressions of the particle update speed and the particle update position are:
Figure FDA0003319791830000023
Figure FDA0003319791830000024
in the above formula, the first and second carbon atoms are,
Figure FDA0003319791830000025
is the particle velocity at time t +1, ω is the weighting coefficient,
Figure FDA0003319791830000026
is the particle velocity at time t +1, c1And c2Is a learning factor, rand is [0,1 ]]A random number in between, and a random number,
Figure FDA0003319791830000027
for the optimal position that the particle itself has experienced,
Figure FDA0003319791830000028
is the position of the particle at time t,
Figure FDA0003319791830000029
for the optimal position that the particle population has experienced,
Figure FDA00033197918300000210
is the position of the particle at time t +1,
Figure FDA00033197918300000211
and
Figure FDA00033197918300000212
continuously updating in the iterative process, and finally outputting the optimal solution
Figure FDA0003319791830000031
The calculation function expression of the weighting coefficient ω is shown as follows:
Figure FDA0003319791830000032
in the above formula, ωmaxIs the maximum value of the weighting coefficient, ωminIs the minimum value of the weighting coefficient, favgIs the average value of the fitness of the particles,fmaxis the maximum value of the fitness of the particle, fminIs the minimum value of the particle fitness, and f is the current particle fitness.
8. The method for adjusting power purchase structure according to claim 7, wherein the step 3) of inputting the processed data into a machine learning classification model which is constructed in advance and trained to predict the power purchase cost comprises: for sample set { (x)i,yi)|i＝1,2,…,N}，xiRepresenting a sample, xie.Rn, where Rn represents a set of real numbers, N represents the number of samples, yiE { -1,1} represents a sample label, and an optimal binary hyper-plane is established as shown in the following formula:
WΦ(x)+b＝0
in the above formula, W is a normal vector of the hyperplane, phi (x) is nonlinear transformation, and b is a constant term of the hyperplane;
through the optimal secondary classification hyperplane, the nonlinear feature vector is mapped into a plane space, the classification problem can be converted into an optimal secondary classification hyperplane optimization problem, and according to a risk minimization principle, the linear indivisible problem is converted into a linear constraint optimization problem:
Figure FDA0003319791830000033
in the above formula, L represents a Lagrangian function, N is the number of points, aiIs a Lagrangian factor, ajIs Lagrangian factor, yiThe category of the ith point is marked as 1 or-1, yjIs the category to which the jth point belongs, xiIs the feature vector of the ith point, xjThe feature vector of the j point; and the machine learning classification model adopts an SVM prediction model, and a model decision function of the SVM prediction model is selected as a Gaussian radial basis kernel function:
Figure FDA0003319791830000034
in the above equation, K (X, Y) is a gaussian radial basis function, X, Y are two vectors, | X-Y | is the distance between vectors X, Y, σ is a constant and σ ≠ 0.
9. A power purchase structure adjustment system comprising a microprocessor and a memory connected to each other, wherein the microprocessor is programmed or configured to perform the steps of the power purchase structure adjustment method according to any one of claims 1 to 8, or the memory stores a computer program programmed or configured to perform the power purchase structure adjustment method according to any one of claims 1 to 8.
10. A computer-readable storage medium having stored thereon a computer program programmed or configured to perform the method of adjusting an electricity purchasing architecture according to any one of claims 1 to 8.
CN202111242725.8A 2021-10-25 2021-10-25 Electricity purchasing structure adjusting method, system and medium Pending CN113988919A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202111242725.8A CN113988919A (en) 2021-10-25 2021-10-25 Electricity purchasing structure adjusting method, system and medium

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202111242725.8A CN113988919A (en) 2021-10-25 2021-10-25 Electricity purchasing structure adjusting method, system and medium

Publications (1)

Publication Number Publication Date
CN113988919A true CN113988919A (en) 2022-01-28

Family

ID=79741148

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202111242725.8A Pending CN113988919A (en) 2021-10-25 2021-10-25 Electricity purchasing structure adjusting method, system and medium

Country Status (1)

Country Link
CN (1) CN113988919A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114511360A (en) * 2022-04-20 2022-05-17 华能江西能源销售有限责任公司 Electricity purchase management method, system, readable storage medium and computer equipment

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114511360A (en) * 2022-04-20 2022-05-17 华能江西能源销售有限责任公司 Electricity purchase management method, system, readable storage medium and computer equipment
CN114511360B (en) * 2022-04-20 2022-07-08 华能江西能源销售有限责任公司 Electricity purchase management method, system, readable storage medium and computer equipment

Similar Documents

Publication Publication Date Title
Lin et al. Short-term load forecasting based on LSTM networks considering attention mechanism
Imani Electrical load-temperature CNN for residential load forecasting
Zhou et al. Parameter optimization of nonlinear grey Bernoulli model using particle swarm optimization
CN111199016A (en) DTW-based improved K-means daily load curve clustering method
Phyo et al. Electricity load forecasting in Thailand using deep learning models
CN109214449A (en) A kind of electric grid investment needing forecasting method
CN108171379B (en) Power load prediction method
CN110751326B (en) Photovoltaic day-ahead power prediction method and device and storage medium
CN110674993A (en) User load short-term prediction method and device
Wang et al. A deep asymmetric Laplace neural network for deterministic and probabilistic wind power forecasting
Zhang et al. A new solar power output prediction based on hybrid forecast engine and decomposition model
CN113537469B (en) Urban water demand prediction method based on LSTM network and Attention mechanism
CN113255900A (en) Impulse load prediction method considering improved spectral clustering and Bi-LSTM neural network
CN110895772A (en) Electricity sales amount prediction method based on combination of grey correlation analysis and SA-PSO-Elman algorithm
CN114065653A (en) Construction method of power load prediction model and power load prediction method
CN114492191A (en) Heat station equipment residual life evaluation method based on DBN-SVR
Wang et al. Big data analytics for price forecasting in smart grids
Dong et al. Short-term building cooling load prediction model based on DwdAdam-ILSTM algorithm: A case study of a commercial building
CN116187835A (en) Data-driven-based method and system for estimating theoretical line loss interval of transformer area
CN112365056A (en) Electrical load joint prediction method and device, terminal and storage medium
CN113988919A (en) Electricity purchasing structure adjusting method, system and medium
CN114581141A (en) Short-term load prediction method based on feature selection and LSSVR
Wu et al. A structure with density-weighted active learning-based model selection strategy and meteorological analysis for wind speed vector deterministic and probabilistic forecasting
CN112465266A (en) Bus load prediction accuracy analysis method and device and computer equipment
Wang et al. Cloud computing and extreme learning machine for a distributed energy consumption forecasting in equipment-manufacturing enterprises

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