CN111861089A - Comprehensive evaluation method for electric power spot market - Google Patents

Comprehensive evaluation method for electric power spot market Download PDF

Info

Publication number
CN111861089A
CN111861089A CN202010255248.8A CN202010255248A CN111861089A CN 111861089 A CN111861089 A CN 111861089A CN 202010255248 A CN202010255248 A CN 202010255248A CN 111861089 A CN111861089 A CN 111861089A
Authority
CN
China
Prior art keywords
index
market
weight
evaluation
electric power
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
CN202010255248.8A
Other languages
Chinese (zh)
Inventor
吴一峰
张思
黄远平
李洋
陈菁伟
黄剑峰
吕磊炎
李璋
杨靖萍
徐建平
吕勤
胡济恒
侯佳萱
蒋轶澄
虞佳淼
江昕玥
张智
林振智
杨莉
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 Zhejiang Electric Power Co Ltd
Jinhua Power Supply Co of State Grid Zhejiang Electric Power Co Ltd
Original Assignee
State Grid Zhejiang Electric Power Co Ltd
Jinhua Power Supply Co of State Grid Zhejiang 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 Zhejiang Electric Power Co Ltd, Jinhua Power Supply Co of State Grid Zhejiang Electric Power Co Ltd filed Critical State Grid Zhejiang Electric Power Co Ltd
Priority to CN202010255248.8A priority Critical patent/CN111861089A/en
Publication of CN111861089A publication Critical patent/CN111861089A/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
    • G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
    • G06Q10/063—Operations research, analysis or management
    • G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
    • G06Q10/06393—Score-carding, benchmarking or key performance indicator [KPI] analysis
    • 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
    • G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
    • G06Q10/063—Operations research, analysis or management
    • G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
    • 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
    • G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
    • G06Q10/063—Operations research, analysis or management
    • G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
    • G06Q10/06395—Quality analysis or management
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
    • G06Q50/06—Energy or water supply

Landscapes

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

Abstract

The invention discloses a comprehensive evaluation method for a power spot market. Firstly, starting from the principle of the electric power spot market, determining the principle of index selection, including the aspects of comprehensiveness, simplicity, feasibility, comparability, guidance and focus. Meanwhile, based on the consideration of a complex system, a hierarchical index system is selected as a basic framework of an evaluation system. Then, a plurality of groups of weight coefficients are obtained by adopting various weight assignment methods such as an expert scoring method, a correlation coefficient analysis method, an entropy value method based on an analytic hierarchy process and the like, and the comprehensive weight coefficients of all indexes are determined by utilizing a comprehensive weight assignment method based on a game theory. And finally, comprehensively evaluating the market by using a grey correlation degree analysis method in combination with the grey characteristic of the electric power spot market. The method has an important role in scientifically analyzing the operation condition of the electric power market, finding out existing problems and predicting the market development trend, thereby promoting the healthy and orderly development of the electric power spot market.

Description

Comprehensive evaluation method for electric power spot market
Technical Field
The invention relates to the technical field of power markets, in particular to a comprehensive evaluation method for a power spot market.
Background
The electric power spot market is a trading market which carries out real-time trading or quasi real-time trading by taking electric energy as a commodity, and is constructed for introducing market competition in the electric power industry, and the market competition brought by marketization is utilized to guide the optimal configuration of the overall social resources of the electric power spot market so as to obtain greater social benefits.
The improvement of the power market began abroad in the last 90 th century. The power market revolution was first initiated in western europe and subsequently generalized to some utility companies in the united states and across oceans. By 2009, 86.49% of the global population underwent power marketization reform, with the national GDP of power marketization comprehensively occupying 94% of the global GDP. At present, the mature electric power trading markets in the world mainly include electric power markets such as the PJM electric power spot market, the northern european electric power market, the british electric power market, and the australian electric power market in the united states.
The market reformation of electric power in 2002 begins in China, and certain achievements have been achieved in the past 17 years. At present, the southern (started in the Guangdong) electric power spot market is established as a regional electric power spot market in eight spot test areas selected nationwide, and the first electric power spot trade settlement in China is carried out between 5, 15 and 16 days in 2019, so that the construction of a real-time market is promoted. In addition, the test run of the electric power spot market is about to be started in Zhejiang.
However, the construction of the electric power market in China is still in the stage of preliminary exploration, and an effective electric power market evaluation index system is not established yet. Under the new market competition, the electric power market information changes instantly, and the occurrence of poor operation or market crisis in the market can cause serious damage to regions and even the whole electric power industry. Therefore, whether the power generator or the power purchasing party in the market, the market operator or the market supervisor pay attention to the operation condition of the market, and the market health degree needs to be analyzed and evaluated. Therefore, the construction of the electric power spot market evaluation system and the determination of the comprehensive evaluation method thereof are both an important direction of the electric power market reform in China at present and are key discussion objects of the next stage of electric power market construction.
Disclosure of Invention
The invention mainly constructs a power spot market evaluation system based on a centralized market and provides a multi-attribute comprehensive evaluation method based on grey correlation degree analysis of a comprehensive weight coefficient.
The invention is realized by adopting the following technical scheme:
a comprehensive evaluation method for a power spot market comprises the following steps:
s1: selecting a hierarchical index system as a basic framework of an evaluation system from the economic characteristics of an electric power market and the physical characteristics of an electric power system;
S2: obtaining a power spot market evaluation index weight coefficient by adopting an expert scoring method;
s3: calculating a power spot market evaluation index weight coefficient by adopting a correlation coefficient analysis method;
s4: calculating a power spot market evaluation index weight coefficient by adopting an entropy method based on an analytic hierarchy process;
s5: determining the comprehensive weight coefficient of each index by utilizing a comprehensive weight assignment method based on game theory and combining the advantages of various weighting methods;
s6: and (3) performing comprehensive evaluation on the market by using a grey correlation degree analysis method in combination with the grey characteristic of the electric power spot market.
In the above technical solution, further, from the economic characteristics of the power market and the physical characteristics of the power system, S1 selects a hierarchical index system as a basic frame of an evaluation system, and the specific implementation method is as follows:
and a hierarchical power spot market evaluation index system is established according to the principles of comprehensiveness, simplicity, feasibility, comparability, guidance and focus. The market safety level, the market operation level, the market benefit level and the market scheduling level are used as main operation state indexes of the market for evaluation, a complete and scientific four-stage power spot market evaluation system comprising a system layer, a state layer, a structural layer and an index layer is developed and constructed, and a secondary index layer is developed downwards again in some indexes, for example, the voltage fluctuation indexes are developed downwards again into three secondary indexes of positive voltage fluctuation rate, negative voltage fluctuation rate and voltage fluctuation influence time. The established electric power spot market evaluation system combines the national conditions of China to constructively put forward indexes such as local structure indexes, market guarantee information safety level, market settlement indexes and the like of the market at the electricity purchasing side. Table 4 shows the overall framework of the electric power spot market evaluation index system.
TABLE 4 electric power spot market evaluation index system integral framework
Figure RE-GDA0002637073780000021
Figure RE-GDA0002637073780000031
Figure RE-GDA0002637073780000041
Further, in step S2, an expert scoring method is used to obtain the weight coefficient of the power spot market evaluation index, and the specific implementation method is as follows:
the expert scoring method, also known as the delphi method, is the simplest and the most clear method in the subjective weighting method. The method mainly comprises the steps of scoring index weights in an evaluation system by means of experience, knowledge and personal preference of a plurality of experts having high familiarity with an evaluation object, and properly processing scoring results by means of a statistical method to obtain a final weight coefficient. The specific evaluation model is as follows:
(1) establishing a preliminary index weight matrix according to the scoring opinions of experts on each index
Figure RE-GDA0002637073780000042
In the formula: xiRepresenting the assignment scheme of the ith expert to the index weight; omegaijRepresenting the assignment of the ith expert to the jth index weight.
(2) Calculating the weight coefficient average value of each index
Figure RE-GDA0002637073780000043
In the formula:
Figure RE-GDA0002637073780000044
represents the average value of the jth index weight coefficient.
(3) Calculating the difference level of the weight coefficient and the average value of each index
Figure RE-GDA0002637073780000045
In the formula: Δ ωijRepresenting the difference between the i-th expert's assignment to the j-th index weight and the average of all experts' assignments to the j-th index weight.
(4) For the difference Δ ωjAnd (3) inviting a plurality of experts to re-evaluate the jth index with a larger comparison, giving a weight coefficient, repeating the steps (1) to (3) until all the experts in the evaluation system evaluate the jth index and the weight coefficient are assigned within a certain deviation, and recording a primary index weight matrix obtained at the last time:
Figure RE-GDA0002637073780000051
the matrix is used to calculate the average weight assignment of each index and the weight assignment is used as the final index weight, that is, the matrix is used to calculate the average weight assignment of each index
Figure RE-GDA0002637073780000052
In the formula: p is a radical of1Representing a weight coefficient vector obtained by a first class of weighting methods, i.e., expert weighting.
S3, calculating the weight coefficient of the evaluation index of the spot market of the electric power by adopting a correlation coefficient analysis method, wherein the specific implementation method comprises the following steps:
in the actual index system construction, the indexes are inevitably repeated or similar in function, and if a method similar to an expert scoring method is adopted, the weight of the indexes in the aspect is enlarged or reduced, so that the final evaluation result is not ideal. The correlation coefficient analysis rule is to compare the correlation coefficients between indexes to judge the inherent correlation before the indexes. The larger the correlation is, the higher the coverage rate and repetition rate of information among the indexes are, and the smaller the corresponding weight assignment is; conversely, the smaller the correlation is, the lower the coverage rate and the repetition rate of the information among the indexes are, and the larger the corresponding weight assignment is. In statistics, the pearson correlation coefficient is generally used for quantitative representation of the correlation between two variables, and therefore, the pearson correlation coefficient is used herein to characterize the correlation between indexes.
Figure RE-GDA0002637073780000053
In the formula: rho(p,q)Representing a correlation coefficient between the index p and the index q; x is the number ofip，xiqRepresenting the p and q index values in the ith electric power spot market;
Figure RE-GDA0002637073780000061
and
Figure RE-GDA0002637073780000062
represents the average of the p and q indices in the electricity spot market.
When one or more sets of indices in the index values are constant values, the denominator is 0 when the correlation coefficient is calculated, and the calculation is erroneous. The data is subjected to 0.1% of small disturbance, and meanwhile, the sum of index values is guaranteed to be unchanged, so that the problem that the denominator is zero and cannot be calculated can be solved.
A matrix of correlation coefficients can be written:
Figure RE-GDA0002637073780000063
thus, the weight coefficient of each index can be determined as:
Figure RE-GDA0002637073780000064
in the formula: omegajThe weight coefficient representing the j-th index.
The index weight is expressed in the form of a weight vector as:
p2＝[ω1,ω2,L,ωi,L,ωn](i＝1,2,L,n) (44)
in the formula: p is a radical of2The weight coefficient vector obtained by the second weighting method, i.e., the correlation coefficient analysis method, is shown.
S4, calculating the weight coefficient of the evaluation index of the electric power spot market by an entropy method based on an analytic hierarchy process, wherein the specific implementation method comprises the following steps:
the combined weighting method is provided based on an optimization theory, has the characteristic of covering subjective and objective weighting, and can avoid the excessive influence of human factors and sample difference on the importance degree of indexes. An entropy method based on analytic hierarchy process is used to perform the combined weighting of the indicators.
The analytic hierarchy process is a subjective weighting method and has the characteristic of combining qualitative analysis with quantitative analysis to process various decision factors. The basic idea is to determine a judgment matrix, compare every two judgment matrixes to judge the relative importance of each element in the matrix, and then assign evaluation indexes; the entropy rule is an objective weighting method, and index weights are assigned by using information entropy.
In the information theory, entropy is a measure of uncertainty, and can also be used to determine the degree of dispersion of a certain index. In an evaluation system, when the dispersion degree of one index in different electric power spot markets is smaller, the larger the entropy is, the smaller the uncertainty is, the smaller the information which reflects the difference of the different electric power spot markets is carried, and the smaller the effect played in comparison among markets is, the smaller the weight coefficient is; conversely, the smaller the entropy, the larger the weighting factor.
The main idea of the entropy method based on the analytic hierarchy process is to respectively use the traditional analytic hierarchy process and the entropy method to assign the weight coefficient in each layer of analysis, and multiply the two methods to obtain the combined weight assignment in the aspect. The processing method can reflect the influence of human factors and can reflect the change of policies through market evaluation, thereby having guiding significance on the market development trend; on the other hand, the entropy rule can enlarge the difference between different markets to the current evaluation result, and facilitates comparison and perfection between different markets.
Step 1: carrying out weight assignment on each index in the electric power spot market evaluation system by using a traditional analytic hierarchy process to obtain
Figure RE-GDA0002637073780000071
(1) Constructing a hierarchical structure model according to the established electric power spot market evaluation system;
(2) constructing a decision matrix XAHP
The indexes are compared pairwise according to a proportion scale table listed in the table 5, and the results are written into a judgment matrix XAHP。
TABLE 5 analytic hierarchy Process Scale Table
Relationship between factor A and factor B Quantized value
Of equal importance 1
Of slight importance 3
Of greater importance 5
Of strong importance 7
Of extreme importance 9
Intermediate values of two adjacent judgments 2,4,6,8
Figure RE-GDA0002637073780000081
In the formula:
Figure RE-GDA0002637073780000082
and the importance degree judgment result between the ith index and the jth index is shown.
At the same time, the matrix X is judgedAHPThe method also has the following two characteristics:
Figure RE-GDA0002637073780000083
(3) obtaining a judgment matrix XAHPMaximum feature root of
Figure RE-GDA0002637073780000084
And corresponding normalized feature vectors
Figure RE-GDA0002637073780000085
The elements in (b) are ranking weights for the relative importance of higher level criteria and rules at the same level, so the process is also referred to as level ordering.
(4) And (5) checking the consistency. Not all judgment matrixes can carry out hierarchical single sequencing, and when the judgment matrixes determine that the inconsistency has large deviation, the hierarchical single sequencing cannot be carried out. Therefore, it is necessary to perform a consistency check to check for inconsistency in the determination of the determination matrix.
Figure RE-GDA0002637073780000086
In the formula:
Figure RE-GDA0002637073780000087
an ith eigenvalue representing a decision matrix; CIiExpressing the consistency index of the ith eigenvalue of the judgment matrix when the CI is usediWhen 0, there is complete identity, when CIiWhen close to 0, there is satisfactory consistency, when CIiThe greater the inconsistency is; RI represents an average random consistency index of the judgment matrix, and is generally judged using an average random consistency index RI standard value shown in table 6; CIMAXRepresenting the obtained consistency index for the maximum characteristic root; CR represents the check coefficient of the decision matrix, and generally, if CR < 0.1, the decision matrix is considered to pass the consistency check.
TABLE 6 Standard values of average random consistency index RI
Figure RE-GDA0002637073780000091
(5) Improvement of consistency. And if the judgment matrix does not pass the consistency check, re-soliciting the expert opinion for recalculation. However, this would greatly increase the workload that experts may use, and a consistency improvement method is used to adjust the elements of the decision matrix to reduce the consistency to some extent. The steps for consistency improvement are as follows:
1) for judgment matrix XAHPColumn normalization was performed:
Figure RE-GDA0002637073780000092
wherein
Figure RE-GDA0002637073780000093
In the formula: xAHP *Representing a judgment matrix after column normalization;
Figure RE-GDA0002637073780000094
representing the elements in the decision matrix after column normalization.
2) And (3) solving an ordering vector by using a sum-product method:
Figure RE-GDA0002637073780000095
wherein
Figure RE-GDA0002637073780000101
3) Insertion-induced matrix
Figure RE-GDA0002637073780000102
Wherein
Figure RE-GDA0002637073780000103
4) Finding out the position (i, j) of the maximum value element in the induction matrix, and judging the element of the corresponding position in the original judgment matrix
Figure RE-GDA0002637073780000104
Is modified into
Figure RE-GDA0002637073780000105
At the same time will
Figure RE-GDA0002637073780000106
Is modified into
Figure RE-GDA0002637073780000107
5) And (5) carrying out consistency check again, and repeating 1) to 4) if the consistency check is not met until the consistency check is met.
6) Hierarchical ordering and consistency checking. And in each layer, sorting the weights of relative importance of all the factors to a higher layer until the highest layer, namely the target layer is sorted, and simultaneously carrying out consistency check.
Step 2: carrying out weight assignment on each index according to the electric power spot market operation data by using an entropy method to obtain
Figure RE-GDA0002637073780000108
(1) Standardizing each index data, processing according to a formula to obtain a result, and expressing each group of index data as an electric power market operation effect index matrix
Figure RE-GDA0002637073780000111
In the formula: miRepresenting an ith electric power market operation effect index vector; x is the number ofijIndicating the jth index in the ith power market.
(2) Calculating the proportion of each power market under each index
Figure RE-GDA0002637073780000112
In the formula: pijAnd the weight of the jth index in the ith power market in the jth index is represented.
(3) Calculating an entropy value for each index
Figure RE-GDA0002637073780000113
Wherein
Figure RE-GDA0002637073780000114
In the formula: e.g. of the typejEntropy representing the jth index; k is a correction coefficient.
(4) Calculating a weight coefficient of each index
Figure RE-GDA0002637073780000115
In the formula: p is a radical ofjThe weight coefficient representing the j-th index.
And step 3: weighting coefficient obtained by traditional analytic hierarchy process
Figure RE-GDA0002637073780000116
Weight coefficient calculated by entropy method
Figure RE-GDA0002637073780000117
Multiplying to obtain a combined coefficient
Figure RE-GDA0002637073780000118
Namely:
Figure RE-GDA0002637073780000121
and 4, step 4: if there is a higher level of rating hierarchy, the above steps are repeated again.
And 5: and recalculating the weight assignment of all indexes according to the assignment in the previous step, namely:
Figure RE-GDA0002637073780000122
in the formula:
Figure RE-GDA0002637073780000123
the final index weight coefficient is represented by,
Figure RE-GDA0002637073780000124
obtaining a weight coefficient vector p of indexes in a final electric power spot market evaluation system3Expressed as:
Figure RE-GDA0002637073780000125
in the formula: p is a radical of3And the weight coefficient vector obtained by adopting a third weighting method, namely an entropy method based on an analytic hierarchy process is shown.
S5, determining the comprehensive weight coefficient of each index by using a comprehensive weight assignment method based on the game theory and combining the advantages of various weighting methods, wherein the specific implementation method is as follows:
step 1: calculating a target weight vector ωres:
Figure RE-GDA0002637073780000126
In the formula: omegaiRepresenting weight vectors obtained under the method in the ith under the k weight coefficient determination methods; alpha is alpha iAnd the linear combination coefficient represents the influence degree of the weight vector obtained under the method in the i in the target weight vector.
Step 2: representing target weight vector set as vector set in a centralized manner
Figure RE-GDA0002637073780000127
And step 3: finding the optimal alphaiThe deviation between the set of target weight vectors and the respective weight vectors is minimized, i.e.:
Figure RE-GDA0002637073780000131
the conditions for the optimized first derivative according to its differential properties are:
Figure RE-GDA0002637073780000132
and 4, step 4: from the solved optimal alphaiRepresents the final integrated weight vector ωop。
S6: the method is characterized in that the grey characteristic of the electric power spot market is combined, a grey correlation degree analysis method is used for comprehensively evaluating the market, and the specific implementation method is as follows:
the three weighting factors are defined as follows according to the reason for generating the weighting factors:
(1) basic weight coefficient
The basic weight coefficient refers to a weight coefficient obtained by using an expert scoring method. The method is a result obtained according to the knowledge and experience of experts, reflects the academic world and the professional opinion of the engineering on the problem to a certain extent, has very important reference significance, and can be used as the basic assignment of the weight coefficient.
(2) Possible weight coefficient
The possible weight coefficient refers to a weight coefficient obtained by an objective weighting method. The objective weighting method assigns the indexes according to the difference between data, namely, the objective weighting method indicates the indexes which need to focus on in a specific calculation example for a decision maker in specific application. And are therefore referred to as possible weight coefficients.
(3) Reference weight coefficient
The reference weight coefficient refers to a result obtained by a combination weight assignment method. The method simply considers the advantages and the disadvantages of the subjective and objective valuation method, has certain optimization, can be used as the final weight coefficient of some simpler index systems, and is called as a reference weight coefficient. However, when the method is applied to a huge and complex evaluation system like an electric power spot market evaluation system, the method is simple and thin, and the reliability is not high enough.
It can be seen that, in the foregoing, inconsistent relationships exist between weight coefficients obtained by using different methods, and in order to coordinate weights between the methods and obtain an equilibrium result, so that the weight coefficients are more scientific, comprehensive and objective, and can be applied to complex and large evaluation systems, a combination weight determination method based on a game theory is adopted herein. According to the method, different weight coefficients are integrated according to the principle of game theory, inconsistency among different methods is coordinated, and final more reasonable weight coefficient assignment is obtained.
The weight-collecting model can be generally divided into the following three types: game staging models, team staging models, and group staging models. The combined weight determining method based on the game theory adopts a game aggregation model, and the basic idea is to find a combined weight coefficient vector from different weight coefficient vectors, so that the sum of the distances between the vector and other vectors is the minimum value, namely the deviation between the comprehensive weight and the weights obtained by various methods is the minimum, and the essence is a multi-user optimization problem.
Step 1: standardizing each index data, processing according to a formula to obtain a result, and expressing each group of index data as the following power market operation effect evaluation matrix
Figure RE-GDA0002637073780000141
In the formula: ma0Represents the evaluation vector of the electric power market operation effect in the theoretical operation state, st0jThe j-th index value under the theoretical operation state is generally regarded as 1, Ma by defaultiRepresents the ith electric power market operation effect evaluation vector, stijIndicating the value of the jth index in the ith power market.
Step 2: the evaluation matrix of the operation effect of the electric power market is obtained by multiplying the comprehensive index weight coefficient obtained by the game aggregation model, namely the evaluation matrix of the comprehensive operation effect of the electric power market is obtained
Figure RE-GDA0002637073780000151
In the formula: CMa0Represents the comprehensive operation effect evaluation vector, Cst, of the power market in the theoretical operation state0jRepresents the j-th index comprehensive evaluation value, CMa, in the theoretical operating stateiRepresents the comprehensive operation effect evaluation vector, Cst, of the ith power marketijIndicates the j-th row index comprehensive evaluation value, omega, in the ith power marketjAnd the comprehensive weight coefficient of the j index.
And step 3: matrix for evaluating operation effect of power market by calculating difference array
Figure RE-GDA0002637073780000152
Wherein
ΔCstij＝|Cst0j-Cstij|(i＝1,2,L,m；j＝1,2,L,n) (68)
In the formula: delta CMaiAnd the difference vector represents the comprehensive operation effect evaluation vector of the ith power market and the operation effect evaluation vector in the theoretical operation state.
And 4, step 4: solving the grey correlation coefficient of each index in each market and the corresponding index in the theoretical running state
Figure RE-GDA0002637073780000161
In the formula, the value of delta m is the minimum value in the array matrix of the difference of the power market operation effect evaluation matrix
Figure RE-GDA0002637073780000162
Delta M is the maximum value in the difference array matrix of the power market operation effect evaluation matrix, namely
Figure RE-GDA0002637073780000163
Rho is a resolution coefficient and has a value range of [0, 1%]Influence the Gray correlation coefficient xiijBut has no effect on the final correlation coefficient ordering, typically taking an empirical value of 0.618.
And 5: calculating the association degree of each market and the market under the theoretical operation state
Figure RE-GDA0002637073780000164
Step 6: the obtained correlation degree gammaiSorting from big to small to obtain the sorting from good to bad about the operation condition of the i power markets and obtain a relative score value, namely a numerical value gamma of the relevancei。
The technical scheme provided by the invention has the beneficial effects that:
in order to enable the weight coefficients to be more scientific and reasonable, the invention utilizes a comprehensive weight coefficient determining method based on game theory to carry out weight coefficient reacquisition on three groups of weight coefficients obtained by an objective weighting method and a combined weighting method. In the method, a subjective value assigning method brings a function of artificially increasing the weight of a certain index to reflect policy guidance, an objective value assigning method brings a function of screening data to find data with the maximum difference for comparison, and a combined value assigning method brings a weight coefficient with certain scientific rationality as a reference. Therefore, the comprehensive weight assignment method provided by the invention has great advantages compared with other single assignment methods or simple combination assignment methods. On the basis, the invention adopts a grey correlation degree analysis method suitable for a grey system to carry out comprehensive evaluation on the electric power spot market. The method mainly uses the existing data and the data of the theoretical optimal state to compare the similarity or the dissimilarity degree, thereby obtaining the relative quality judgment between the electric power spot markets, being beneficial to a market manager to know the running state of the market and analyze and solve the existing problems.
Drawings
FIG. 1 is a schematic overall flow diagram of the present invention.
FIG. 2 is an index system basic framework of the present invention.
FIG. 3 is a comparison of index weights under the method for determining all the weight coefficients of the medium-long term transaction part of the present invention.
Detailed Description
The drawings are for illustrative purposes only and are not to be construed as limiting the patent; it will be understood by those skilled in the art that certain well-known structures in the drawings and descriptions thereof may be omitted. The positional relationships depicted in the drawings are for illustrative purposes only and are not to be construed as limiting the present patent.
The invention relates to a comprehensive evaluation method of a power spot market, which comprises the following implementation processes of:
s1, starting from the economic characteristics of the power market and the physical characteristics of the power system, selecting a hierarchical index system as a basic frame of an evaluation system, and specifically realizing the method as follows:
and a hierarchical power spot market evaluation index system is established according to the principles of comprehensiveness, simplicity, feasibility, comparability, guidance and focus. The market safety level, the market operation level, the market benefit level and the market scheduling level are used as main operation state indexes of the market for evaluation, a complete and scientific four-stage power spot market evaluation system comprising a system layer, a state layer, a structural layer and an index layer is developed and constructed, and a secondary index layer is developed downwards again in some indexes, for example, the voltage fluctuation indexes are developed downwards again into three secondary indexes of positive voltage fluctuation rate, negative voltage fluctuation rate and voltage fluctuation influence time. The established electric power spot market evaluation system combines the national conditions of China to constructively put forward indexes such as local structure indexes, market guarantee information safety level, market settlement indexes and the like of the market at the electricity purchasing side. Table 7 shows the overall framework of the electric spot market evaluation index system.
Table 7 electric power spot market evaluation index system integral frame
Figure RE-GDA0002637073780000171
Figure RE-GDA0002637073780000181
S2, obtaining the weight coefficient of the evaluation index of the spot power market by adopting an expert scoring method, wherein the specific implementation method is as follows:
the expert scoring method, also known as the delphi method, is the simplest and the most clear method in the subjective weighting method. The method mainly comprises the steps of scoring index weights in an evaluation system by means of experience, knowledge and personal preference of a plurality of experts having high familiarity with an evaluation object, and properly processing scoring results by means of a statistical method to obtain a final weight coefficient. The specific evaluation model is as follows:
(1) establishing a preliminary index weight matrix according to the scoring opinions of experts on each index
Figure RE-GDA0002637073780000191
In the formula: xiRepresenting the assignment scheme of the ith expert to the index weight; omegaijRepresenting the assignment of the ith expert to the jth index weight.
(2) Calculating the weight coefficient average value of each index
Figure RE-GDA0002637073780000192
In the formula:
Figure RE-GDA0002637073780000193
represents the average value of the jth index weight coefficient.
(3) Calculating the difference level of the weight coefficient and the average value of each index
Figure RE-GDA0002637073780000194
In the formula: Δ ωijRepresenting the difference between the i-th expert's assignment to the j-th index weight and the average of all experts' assignments to the j-th index weight.
(4) For the difference Δ ωjThe j-th index which is relatively large needs to invite a plurality of experts again to reevaluate the j-th index, a weight coefficient is given, and the steps (1) to (3) are repeated until each index in the evaluation systemThe evaluation of all experts of the index on the index and the assignment of the weight coefficient are within a certain deviation, and a primary index weight matrix obtained at the last time is recorded as follows:
Figure RE-GDA0002637073780000201
the matrix is used to calculate the average weight assignment of each index and the weight assignment is used as the final index weight, that is, the matrix is used to calculate the average weight assignment of each index
Figure RE-GDA0002637073780000202
In the formula: p is a radical of1Representing a weight coefficient vector obtained by a first class of weighting methods, i.e., expert weighting.
S3, calculating the weight coefficient of the evaluation index of the spot market of the electric power by adopting a correlation coefficient analysis method, wherein the specific implementation method comprises the following steps:
in the actual index system construction, the indexes are inevitably repeated or similar in function, and if a method similar to an expert scoring method is adopted, the weight of the indexes in the aspect is enlarged or reduced, so that the final evaluation result is not ideal. The correlation coefficient analysis rule is to compare the correlation coefficients between indexes to judge the inherent correlation before the indexes. The larger the correlation is, the higher the coverage rate and repetition rate of information among the indexes are, and the smaller the corresponding weight assignment is; conversely, the smaller the correlation is, the lower the coverage rate and the repetition rate of the information among the indexes are, and the larger the corresponding weight assignment is. In statistics, the pearson correlation coefficient is generally used for quantitative representation of the correlation between two variables, and therefore, the pearson correlation coefficient is used herein to characterize the correlation between indexes.
Figure RE-GDA0002637073780000203
In the formula: rho(p,q)Representing a correlation coefficient between the index p and the index q; x is the number ofip，xiqRepresents the ith powerThe p and q index values in the spot market;
Figure RE-GDA0002637073780000204
and
Figure RE-GDA0002637073780000205
represents the average of the p and q indices in the electricity spot market.
When one or more sets of indices in the index values are constant values, the denominator is 0 when the correlation coefficient is calculated, and the calculation is erroneous. The data is subjected to 0.1% of small disturbance, and meanwhile, the sum of index values is guaranteed to be unchanged, so that the problem that the denominator is zero and cannot be calculated can be solved.
A matrix of correlation coefficients can be written:
Figure RE-GDA0002637073780000211
thus, the weight coefficient of each index can be determined as:
Figure RE-GDA0002637073780000212
in the formula: omegajThe weight coefficient representing the j-th index.
The index weight is expressed in the form of a weight vector as:
p2＝[ω1,ω2,L,ωi,L,ωn](i＝1,2,L,n) (79)
in the formula: p is a radical of2The weight coefficient vector obtained by the second weighting method, i.e., the correlation coefficient analysis method, is shown.
S4, calculating the weight coefficient of the evaluation index of the electric power spot market by an entropy method based on an analytic hierarchy process, wherein the specific implementation method comprises the following steps:
the combined weighting method is provided based on an optimization theory, has the characteristic of covering subjective and objective weighting, and can avoid the excessive influence of human factors and sample difference on the importance degree of indexes. An entropy method based on analytic hierarchy process is used to perform the combined weighting of the indicators.
The analytic hierarchy process is a subjective weighting method and has the characteristic of combining qualitative analysis with quantitative analysis to process various decision factors. The basic idea is to determine a judgment matrix, compare every two judgment matrixes to judge the relative importance of each element in the matrix, and then assign evaluation indexes; the entropy rule is an objective weighting method, and index weights are assigned by using information entropy.
In the information theory, entropy is a measure of uncertainty, and can also be used to determine the degree of dispersion of a certain index. In an evaluation system, when the dispersion degree of one index in different electric power spot markets is smaller, the larger the entropy is, the smaller the uncertainty is, the smaller the information which reflects the difference of the different electric power spot markets is carried, and the smaller the effect played in comparison among markets is, the smaller the weight coefficient is; conversely, the smaller the entropy, the larger the weighting factor.
The main idea of the entropy method based on the analytic hierarchy process is to respectively use the traditional analytic hierarchy process and the entropy method to assign the weight coefficient in each layer of analysis, and multiply the two methods to obtain the combined weight assignment in the aspect. The processing method can reflect the influence of human factors and can reflect the change of policies through market evaluation, thereby having guiding significance on the market development trend; on the other hand, the entropy rule can enlarge the difference between different markets to the current evaluation result, and facilitates comparison and perfection between different markets.
Step 1: carrying out weight assignment on each index in the electric power spot market evaluation system by using a traditional analytic hierarchy process to obtain
Figure RE-GDA0002637073780000221
(1) Constructing a hierarchical structure model according to the established electric power spot market evaluation system;
(2) constructing a decision matrix XAHP
Pairwise comparisons between indices were made according to the scale of the scales listed in Table 8, and the results were writtenMaking a judgment matrix XAHP。
TABLE 8 analytic hierarchy process scaling Table
Relationship between factor A and factor B Quantized value
Of equal importance 1
Of slight importance 3
Of greater importance 5
Of strong importance 7
Of extreme importance 9
Intermediate values of two adjacent judgments 2,4,6,8
Figure RE-GDA0002637073780000222
In the formula:
Figure RE-GDA0002637073780000231
and the importance degree judgment result between the ith index and the jth index is shown.
At the same time, the matrix X is judgedAHPThe method also has the following two characteristics:
Figure RE-GDA0002637073780000232
(3) obtaining a judgment matrix XAHPMaximum feature root of
Figure RE-GDA0002637073780000233
And corresponding normalized feature vectors
Figure RE-GDA0002637073780000234
The elements in (b) are ranking weights for the relative importance of higher level criteria and rules at the same level, so the process is also referred to as level ordering.
(4) And (5) checking the consistency. Not all judgment matrixes can carry out hierarchical single sequencing, and when the judgment matrixes determine that the inconsistency has large deviation, the hierarchical single sequencing cannot be carried out. Therefore, it is necessary to perform a consistency check to check for inconsistency in the determination of the determination matrix.
Figure RE-GDA0002637073780000235
In the formula:
Figure RE-GDA0002637073780000236
an ith eigenvalue representing a decision matrix; CIiExpressing the consistency index of the ith eigenvalue of the judgment matrix when the CI is usediWhen 0, there is complete identity, when CIiWhen close to 0, there is satisfactory consistency, when CIiThe greater the inconsistency is; RI represents an average random consistency index of the judgment matrix, and is generally judged using an average random consistency index RI standard value shown in table 9; CIMAXRepresenting the obtained consistency index for the maximum characteristic root; CR represents the check coefficient of the decision matrix, and generally, if CR < 0.1, the decision matrix is considered to pass the consistency check.
TABLE 9 average random uniformity index RI standard value
Figure RE-GDA0002637073780000237
(5) Improvement of consistency. And if the judgment matrix does not pass the consistency check, re-soliciting the expert opinion for recalculation. However, this would greatly increase the workload that experts may use, and a consistency improvement method is used to adjust the elements of the decision matrix to reduce the consistency to some extent. The steps for consistency improvement are as follows:
1) for judgment matrix XAHPColumn normalization was performed:
Figure RE-GDA0002637073780000241
wherein
Figure RE-GDA0002637073780000242
In the formula: xAHP *Representing a judgment matrix after column normalization;
Figure RE-GDA0002637073780000243
representing the elements in the decision matrix after column normalization.
2) And (3) solving an ordering vector by using a sum-product method:
Figure RE-GDA0002637073780000244
wherein
Figure RE-GDA0002637073780000245
3) Insertion-induced matrix
Figure RE-GDA0002637073780000251
Wherein
Figure RE-GDA0002637073780000252
4) Finding out the position (i, j) of the maximum value element in the induction matrix, and judging the element of the corresponding position in the original judgment matrix
Figure RE-GDA0002637073780000253
Is modified into
Figure RE-GDA0002637073780000254
At the same time will
Figure RE-GDA0002637073780000255
Is modified into
Figure RE-GDA0002637073780000256
5) And (5) carrying out consistency check again, and repeating 1) to 4) if the consistency check is not met until the consistency check is met.
6) Hierarchical ordering and consistency checking. And in each layer, sorting the weights of relative importance of all the factors to a higher layer until the highest layer, namely the target layer is sorted, and simultaneously carrying out consistency check.
Step 2: carrying out weight assignment on each index according to the electric power spot market operation data by using an entropy method to obtain
Figure RE-GDA0002637073780000258
(1) Standardizing each index data, processing according to a formula to obtain a result, and expressing each group of index data as an electric power market operation effect index matrix
Figure RE-GDA0002637073780000257
In the formula: miRepresenting an ith electric power market operation effect index vector; x is the number ofijRepresents the ithJ-th index in the electricity market.
(2) Calculating the proportion of each power market under each index
Figure RE-GDA0002637073780000261
In the formula: pijAnd the weight of the jth index in the ith power market in the jth index is represented.
(3) Calculating an entropy value for each index
Figure RE-GDA0002637073780000262
Wherein
Figure RE-GDA0002637073780000263
In the formula: e.g. of the typejEntropy representing the jth index; k is a correction coefficient.
(4) Calculating a weight coefficient of each index
Figure RE-GDA0002637073780000264
In the formula: p is a radical ofjThe weight coefficient representing the j-th index.
And step 3: weighting coefficient obtained by traditional analytic hierarchy process
Figure RE-GDA0002637073780000265
Weight coefficient calculated by entropy method
Figure RE-GDA0002637073780000266
Multiplying to obtain a combined coefficient
Figure RE-GDA0002637073780000267
Namely:
Figure RE-GDA0002637073780000268
and 4, step 4: if there is a higher level of rating hierarchy, the above steps are repeated again.
And 5: and recalculating the weight assignment of all indexes according to the assignment in the previous step, namely:
Figure RE-GDA0002637073780000269
in the formula:
Figure RE-GDA0002637073780000271
the final index weight coefficient is represented by,
Figure RE-GDA0002637073780000272
obtaining a weight coefficient vector p of indexes in a final electric power spot market evaluation system3Expressed as:
Figure RE-GDA0002637073780000273
in the formula: p is a radical of3And the weight coefficient vector obtained by adopting a third weighting method, namely an entropy method based on an analytic hierarchy process is shown.
S5: the comprehensive weight assignment method based on the game theory is utilized, the advantages of various weighting methods are combined, the comprehensive weight coefficient of each index is determined, and the specific implementation method is as follows:
step 1: calculating a target weight vector ωres:
Figure RE-GDA0002637073780000274
In the formula: omegaiRepresenting weight vectors obtained under the method in the ith under the k weight coefficient determination methods; alpha is alpha iAnd the linear combination coefficient represents the influence degree of the weight vector obtained under the method in the i in the target weight vector.
Step 2: representing target weight vector set as vector set in a centralized manner
Figure RE-GDA0002637073780000275
And step 3: finding the optimal alphaiThe deviation between the set of target weight vectors and the respective weight vectors is minimized, i.e.:
Figure RE-GDA0002637073780000276
the conditions for the optimized first derivative according to its differential properties are:
Figure RE-GDA0002637073780000277
and 4, step 4: from the solved optimal alphaiRepresents the final integrated weight vector ωop。
S6: the method is characterized in that the grey characteristic of the electric power spot market is combined, a grey correlation degree analysis method is used for comprehensively evaluating the market, and the specific implementation method is as follows:
the three weighting factors are defined as follows according to the reason for generating the weighting factors:
(1) basic weight coefficient
The basic weight coefficient refers to a weight coefficient obtained by using an expert scoring method. The method is a result obtained according to the knowledge and experience of experts, reflects the academic world and the professional opinion of the engineering on the problem to a certain extent, has very important reference significance, and can be used as the basic assignment of the weight coefficient.
(2) Possible weight coefficient
The possible weight coefficient refers to a weight coefficient obtained by an objective weighting method. The objective weighting method assigns the indexes according to the difference between data, namely, the objective weighting method indicates the indexes which need to focus on in a specific calculation example for a decision maker in specific application. And are therefore referred to as possible weight coefficients.
(3) Reference weight coefficient
The reference weight coefficient refers to a result obtained by a combination weight assignment method. The method simply considers the advantages and the disadvantages of the subjective and objective valuation method, has certain optimization, can be used as the final weight coefficient of some simpler index systems, and is called as a reference weight coefficient. However, when the method is applied to a huge and complex evaluation system like an electric power spot market evaluation system, the method is simple and thin, and the reliability is not high enough.
It can be seen that, in the foregoing, inconsistent relationships exist between weight coefficients obtained by using different methods, and in order to coordinate weights between the methods and obtain an equilibrium result, so that the weight coefficients are more scientific, comprehensive and objective, and can be applied to complex and large evaluation systems, a combination weight determination method based on a game theory is adopted herein. According to the method, different weight coefficients are integrated according to the principle of game theory, inconsistency among different methods is coordinated, and final more reasonable weight coefficient assignment is obtained.
The weight-collecting model can be generally divided into the following three types: game staging models, team staging models, and group staging models. The combined weight determining method based on the game theory adopts a game aggregation model, and the basic idea is to find a combined weight coefficient vector from different weight coefficient vectors, so that the sum of the distances between the vector and other vectors is the minimum value, namely the deviation between the comprehensive weight and the weights obtained by various methods is the minimum, and the essence is a multi-user optimization problem.
Step 1: standardizing each index data, processing according to a formula to obtain a result, and expressing each group of index data as the following power market operation effect evaluation matrix
Figure RE-GDA0002637073780000291
In the formula: ma0Represents the evaluation vector of the electric power market operation effect in the theoretical operation state, st0jThe j-th index value under the theoretical operation state is generally regarded as 1, Ma by defaultiEvaluation vector for representing operation effect of ith electric power market，stijIndicating the value of the jth index in the ith power market.
Step 2: the evaluation matrix of the operation effect of the electric power market is obtained by multiplying the comprehensive index weight coefficient obtained by the game aggregation model, namely the evaluation matrix of the comprehensive operation effect of the electric power market is obtained
Figure RE-GDA0002637073780000292
In the formula: CMa0Represents the comprehensive operation effect evaluation vector, Cst, of the power market in the theoretical operation state0jRepresents the j-th index comprehensive evaluation value, CMa, in the theoretical operating stateiRepresents the comprehensive operation effect evaluation vector, Cst, of the ith power marketijIndicates the j-th row index comprehensive evaluation value, omega, in the ith power marketjAnd the comprehensive weight coefficient of the j index.
And step 3: matrix for evaluating operation effect of power market by calculating difference array
Figure RE-GDA0002637073780000301
Wherein
ΔCstij＝|Cst0j-Cstij|(i＝1,2,L,m；j＝1,2,L,n) (103)
In the formula: delta CMaiAnd the difference vector represents the comprehensive operation effect evaluation vector of the ith power market and the operation effect evaluation vector in the theoretical operation state.
And 4, step 4: solving the grey correlation coefficient of each index in each market and the corresponding index in the theoretical running state
Figure RE-GDA0002637073780000302
In the formula, the value of delta m is the minimum value in the array matrix of the difference of the power market operation effect evaluation matrix
Figure RE-GDA0002637073780000303
Δ M is the electric marketThe maximum value in the running effect evaluation matrix difference array matrix is
Figure RE-GDA0002637073780000304
Rho is a resolution coefficient and has a value range of [0, 1%]Influence the Gray correlation coefficient xiijBut has no effect on the final correlation coefficient ordering, typically taking an empirical value of 0.618.
And 5: calculating the association degree of each market and the market under the theoretical operation state
Figure RE-GDA0002637073780000305
Step 6: the obtained correlation degree gammaiSorting from big to small to obtain the sorting from good to bad about the operation condition of the i power markets and obtain a relative score value, namely a numerical value gamma of the relevancei。
The following description is given in conjunction with the examples
Part of indexes suitable for medium and long term transaction are screened in the electric power spot market evaluation system to form a new index system, the number of the indexes in the index system is small, and the related aspects of the indexes are more, so that the indexes are not graded any more, and specific indexes and data are shown in a table 10:
TABLE 10 original data of middle and long term trading parts of Guangdong electric power market and Yunnan electric power market
Figure RE-GDA0002637073780000311
The data were normalized to give the results shown in table 11:
TABLE 11 part of standardized data for long-term transactions in the Guangdong electric power market and the Yunnan electric power market
Figure RE-GDA0002637073780000312
Figure RE-GDA0002637073780000321
Referring to professors and opinions of rule makers in the aspect of the power market, obtaining each index weight coefficient vector by using an expert scoring method as follows:
p1＝[0.12,0.08,0.15,0.19,0.04,0.12,0.04,0.04,0.22]
the weight coefficient vector of each index obtained by using the correlation coefficient method is as follows:
p2＝[0.10,0.12,0.10,0.11,0.10,0.11,0.12,0.12,0.12]
obtaining each index weight coefficient vector by using an entropy method based on an analytic hierarchy process as follows:
p3＝[0.07,0.01,0.07,0.07,0.00,0.12,0.03,0.00,0.62]
the final comprehensive weight coefficient vector obtained by using the comprehensive index weight determination method based on the game theory is as follows:
pres＝[0.12,0.08,0.14,0.18,0.04,0.12,0.05,0.04,0.23]
comparing the weight coefficients obtained by different weight assignment methods as shown in fig. 3, it can be seen from the figure that the weight coefficients given by the correlation coefficient analysis method are relatively consistent in size, which indicates that no critical index data needing to be specially noted appears in the group of data; entropy method based on analytic hierarchy process for market openness u2Installed capacity growth rate u5The rate of increase in demand u8And growth rate u of the number of market members9The assignment of (a) is more extreme, for u2Weight is assigned to 0.01, for u5And u8The weight is assigned to 0, for u9The weight is assigned to 0.62, which is far above the average level, the data is reviewed and the matrix found is judged because u is 2、u5And u8The importance degree in the judgment matrix is low, so that the weight coefficient in the analytic hierarchy process is low; meanwhile, the difference between the data is small, so that the weight coefficient of the entropy-value method in the weight assignment is small, and the combined weight coefficient of the entropy-value method is small and even zero. And u9Then, in contrast, the weights in the analytic hierarchy process and the entropy methodThe assignments are all large, resulting in far-above-average combining weights. Although it is reasonable from the viewpoint of weight coefficient assignment, in an actual evaluation system, such weight coefficient assignment is too biased, and a case where the market evaluation score is too low occurs, which is not practical, and this is also a limitation of the entropy method based on the analytic hierarchy process. The combined weight determination method based on the game theory can make up the deficiency in the aspect, and as can be clearly seen from fig. 3, the overall distribution of the weight coefficients is more reasonable, and the entropy method based on the analytic hierarchy process does not appear on u9The case of assigning a weight coefficient of far 0.64 super-average level is better for practical application of the comprehensive weight determination method based on game theory. Meanwhile, the result is close to the result of the expert scoring method, which shows that the result can be agreed by the expert and is very reasonable.
And comparing and calculating the running conditions of the middle and long-term transaction parts of the Guangdong power market and the Yunnan power market with the theoretical optimal running condition by utilizing the grey correlation degree to obtain the following results:
table 12 analysis results of grey correlation degree of middle and long term trading parts in Guangdong electric power market and Yunnan electric power market
Figure RE-GDA0002637073780000331
The grey correlation degree can be obtained, the middle-term and long-term trading situation of the Yunnan electric power market is slightly superior to that of the Guangdong electric power market, and actually, the construction achievement of the Yunnan electric power market is also highly evaluated in China and is consistent with the analysis result of the method. The Guangdong is one of the provinces of the first batch of construction of the electric power market in China, and is the province of the forefront construction of the electric power market in China so far, and the real-time transaction test operation of the electric power spot market is started in the electric power market in the present stage, and is superior to the Yunnan electric power market constructed in the market at present. The result that the running condition of the long-term transaction part in the Guangdong electric power market is slightly lower than that of the Yunnan electric power market shows that the experience training obtained in the construction process of the Guangdong electric power market is successfully popularized, and the construction in the last stage has obvious results; meanwhile, the determination of the next-stage day-ahead market construction in the Yunnan electric power market is correct, and the practical situation and the requirement of the market are met.
The electric power market data of Yunnan is analyzed, and the Yunnan electric power market has distinctive characteristics. The new energy ratio of the Yunnan electric power market is higher, the market opening degree is higher, the electricity price is lower, but the electricity price fluctuation condition is larger. The energy structure of Yunnan is mainly composed of hydroelectric power, the percentage of which is more than 69%, and the Yunnan has larger development and utilization on water resources. Because Yunnan belongs to the season wind climate, has obvious seasonal dry and wet characteristics, the electricity generation amount is too large in the flood season in summer, so that the electricity price is reduced, and the electricity price is increased in the drought season in winter because the electricity generation amount is reduced. The electricity price of one year has obvious seasonal characteristics, and the fluctuation of the electricity price is large when the electricity price is as low as 0.11 yuan/kilowatt hour in summer and as high as 0.22 yuan/kilowatt hour in winter. Meanwhile, in summer, the hydropower station takes the measures of 'water abandoning and price keeping', which causes a great deal of waste of resources. Therefore, how to guarantee larger demand consumption in summer flood season is one of the key points of the next step of the construction of the Yunnan electric power market. Two main aspects can be started: firstly, the openness degree is continuously improved on the basis of the existing high openness degree, the full consumption of the electric quantity in the flood season is guaranteed by means of the export electric quantity, and the electric quantity can be popularized to the whole country for study in the future as a typical case for improving the openness degree in the regional electric power market; and secondly, the development of economy in the region is pulled, the economic level is not too high compared with that of Guangdong province in the east of Yunnan province, a lot of potential spaces are worthy of development, and the power market can stimulate the demand and attract investment through additional subsidies on the power utilization side, so that the economic development of the whole province is driven, and benefits are obtained.

Claims (7)

1. A comprehensive evaluation method for a power spot market is characterized by comprising the following steps:
s1: selecting a hierarchical index system as a basic framework of an evaluation system from the economic characteristics of an electric power market and the physical characteristics of an electric power system;
s2: obtaining a power spot market evaluation index weight coefficient by adopting an expert scoring method;
s3: calculating a power spot market evaluation index weight coefficient by adopting a correlation coefficient analysis method;
s4: calculating a power spot market evaluation index weight coefficient by adopting an entropy method based on an analytic hierarchy process;
s5: determining the comprehensive weight coefficient of each index by utilizing a comprehensive weight assignment method based on game theory and combining the advantages of various weighting methods;
s6: and (3) performing comprehensive evaluation on the market by using a grey correlation degree analysis method in combination with the grey characteristic of the electric power spot market.
2. The comprehensive evaluation method of the electric power spot market according to claim 1, characterized in that: s1, starting from the economic characteristics of the power market and the physical characteristics of the power system, selecting a hierarchical index system as a basic frame of an evaluation system, and specifically realizing the method as follows:
a hierarchical power spot market evaluation index system is established according to the principles of comprehensiveness, simplicity, feasibility, comparability, guidance and focus, four aspects of market safety level, market operation level, market benefit level and market scheduling level are used as main operation state indexes of the market for evaluation, a complete and scientific four-level power spot market evaluation system comprising a system layer, a state layer, a structural layer and an index layer is developed and constructed, and a secondary index layer is developed downwards again in some indexes.
3. The comprehensive evaluation method of the electric power spot market according to claim 2, characterized in that: s2, obtaining the weight coefficient of the evaluation index of the spot power market by adopting an expert scoring method, wherein the specific implementation method is as follows:
the expert scoring method is mainly characterized in that index weights in an evaluation system are scored by means of experience, knowledge and personal preference of a plurality of experts with high familiarity with an evaluation object, and scoring results are appropriately processed by means of a statistical method to obtain a final weight coefficient; the specific evaluation model is as follows:
(1) establishing a preliminary index weight matrix according to the scoring opinions of experts on each index
Figure FDA0002437046300000021
In the formula: xiRepresenting the assignment scheme of the ith expert to the index weight; omegaijRepresenting the assignment of the ith expert to the jth index weight;
(2) calculating the weight coefficient average value of each index
Figure FDA0002437046300000022
In the formula:
Figure FDA0002437046300000023
represents the average value of the jth index weight coefficient;
(3) calculating the difference level of the weight coefficient and the average value of each index
Figure FDA0002437046300000024
In the formula: Δ ωijRepresenting the difference between the ith expert's assignment to the jth index weight and the average of all experts' assignments to the jth index weight;
(4) for the difference Δ ω jAnd (3) inviting a plurality of experts to re-evaluate the jth index with a larger comparison, giving a weight coefficient, repeating the steps (1) to (3) until all the experts in the evaluation system evaluate the jth index and the weight coefficient are assigned within a certain deviation, and recording a primary index weight matrix obtained at the last time:
Figure FDA0002437046300000025
the matrix is used to calculate the average weight assignment of each index and the weight assignment is used as the final index weight, that is, the matrix is used to calculate the average weight assignment of each index
Figure FDA0002437046300000026
In the formula: p is a radical of1Representing a weight coefficient vector obtained by a first class of weighting methods, i.e., expert weighting.
4. The comprehensive evaluation method of the electric power spot market according to claim 3, characterized in that: s3, calculating the weight coefficient of the evaluation index of the spot market of the electric power by adopting a correlation coefficient analysis method, wherein the specific implementation method comprises the following steps:
the correlation coefficient analysis rule judges the inherent correlation before the indexes by comparing the correlation coefficients among the indexes, the larger the correlation is, the higher the coverage rate and the repetition rate of the information among the indexes are, and the smaller the corresponding weight assignment is; on the contrary, the smaller the correlation is, the lower the coverage rate and the repetition rate of the information among the indexes are, and the larger the corresponding weight assignment is; pearson's correlation coefficient is used to characterize the correlation between indices:
Figure FDA0002437046300000031
In the formula: rho(p,q)Representing a correlation coefficient between the index p and the index q; x is the number ofip，xiqRepresenting the p and q index values in the ith electric power spot market;
Figure FDA0002437046300000032
and
Figure FDA0002437046300000033
an average value representing the p-th and q-th indexes in the electric power spot market;
when one or more than one group of indexes in the index values are constant values, the denominator is 0 when the correlation coefficient is calculated, calculation is wrong, 0.1% of micro-disturbance is carried out on data, and meanwhile, the sum of the index values is guaranteed to be unchanged, so that the problem that the denominator is zero and calculation cannot be carried out is solved;
a matrix of correlation coefficients can be written:
Figure FDA0002437046300000034
thus, the weight coefficient of each index can be determined as:
Figure FDA0002437046300000035
in the formula: omegajA weight coefficient representing a j-th index;
the index weight is expressed in the form of a weight vector as:
p2＝[ω1,ω2,…,ωi,…,ωn](i＝1,2,…,n) (9)
in the formula: p is a radical of2The weight coefficient vector obtained by the second weighting method, i.e., the correlation coefficient analysis method, is shown.
5. The comprehensive evaluation method of the electric power spot market according to claim 4, characterized in that: s4, calculating the weight coefficient of the evaluation index of the electric power spot market by an entropy method based on an analytic hierarchy process, wherein the specific implementation method comprises the following steps:
step 1: carrying out weight assignment on each index in the electric power spot market evaluation system by using a traditional analytic hierarchy process to obtain
Figure FDA0002437046300000041
(1) Constructing a hierarchical structure model according to the established electric power spot market evaluation system;
(2) constructing a decision matrix XAHP
Comparing the indexes pairwise according to a set proportion scale table, and writing the result into a judgment matrix XAHP；
Figure FDA0002437046300000042
In the formula:
Figure FDA0002437046300000043
representing the judgment result of the importance degree between the ith index and the jth index;
at the same time, the matrix X is judgedAHPThe method also has the following two characteristics:
Figure FDA0002437046300000044
(3) obtaining a judgment matrix XAHPMaximum feature root of
Figure FDA0002437046300000045
And corresponding normalized feature vectors
Figure FDA0002437046300000046
Figure FDA0002437046300000047
The elements in (1) are the ranking weights of relative importance to higher-level standards and rules at the same level, so the process is also called level list ranking;
(4) consistency check
Not all judgment matrixes can be subjected to level single sequencing, and when the judgment matrixes determine that the inconsistency has large deviation, the level single sequencing cannot be performed, so that consistency check is required to be performed, and the inconsistency of the judgment matrixes is checked:
Figure FDA0002437046300000051
in the formula:
Figure FDA0002437046300000052
an ith eigenvalue representing a decision matrix; CIiExpressing the consistency index of the ith eigenvalue of the judgment matrix when the CI is usediWhen 0, there is complete identity, when CIiWhen close to 0, there is satisfactory consistency, when CIiThe greater the inconsistency is; RI represents the average random consistency index of the judgment matrix, and the average random consistency index RI standard value is generally used for judgment; CI MAXRepresenting the obtained consistency index for the maximum characteristic root; CR represents the check coefficient of the judgment matrix, and generally, if CR is less than 0.1, the judgment matrix is considered to pass the consistency check;
(5) improvement of consistency
Adjusting elements of the judgment matrix by adopting a consistency improvement method to reduce the consistency to a certain extent; the steps for consistency improvement are as follows:
1) for judgment matrix XAHPColumn normalization was performed:
Figure FDA0002437046300000053
wherein
Figure FDA0002437046300000054
In the formula: xAHP *Representing a judgment matrix after column normalization;
Figure FDA0002437046300000061
representing elements in the judgment matrix after column normalization;
2) and (3) solving an ordering vector by using a sum-product method:
Figure FDA0002437046300000062
wherein
Figure FDA0002437046300000063
3) Insertion-induced matrix
Figure FDA0002437046300000064
Wherein
Figure FDA0002437046300000065
4) Finding out the position (i, j) of the maximum value element in the induction matrix, and judging the element of the corresponding position in the original judgment matrix
Figure FDA0002437046300000066
Is modified into
Figure FDA0002437046300000067
At the same time will
Figure FDA0002437046300000068
Is modified into
Figure FDA0002437046300000069
5) Carrying out consistency check again, and if the consistency check is not met, repeating 1) to 4) until the consistency check is met;
6) hierarchical ordering and consistency check
In each layer, sorting the weights of relative importance of all factors to a higher layer until the weights are sorted to the highest layer, namely a target layer, and simultaneously carrying out consistency check;
step 2: carrying out weight assignment on each index according to the electric power spot market operation data by using an entropy method to obtain
Figure FDA0002437046300000071
(1) Standardizing each index data, processing according to a formula to obtain a result, and expressing each group of index data as an electric power market operation effect index matrix
Figure FDA0002437046300000072
In the formula: miRepresenting an ith electric power market operation effect index vector; x is the number ofijRepresenting the jth index in the ith power market;
(2) calculating the proportion of each power market under each index
Figure FDA0002437046300000073
In the formula: pijThe specific gravity of the jth index in the ith power market in the jth index is represented;
(3) calculating an entropy value for each index
Figure FDA0002437046300000074
Wherein
Figure FDA0002437046300000075
In the formula: e.g. of the typejEntropy representing the jth index; k is a correction coefficient;
(4) calculating a weight coefficient of each index
Figure FDA0002437046300000076
In the formula: p is a radical ofjA weight coefficient representing a j-th index;
and step 3: weighting coefficient obtained by traditional analytic hierarchy process
Figure FDA0002437046300000081
Weight coefficient calculated by entropy method
Figure FDA0002437046300000082
Multiplying to obtain a combined coefficient
Figure FDA0002437046300000083
Namely:
Figure FDA0002437046300000084
and 4, step 4: if the evaluation hierarchy with higher level exists, repeating the steps again;
and 5: and recalculating the weight assignment of all indexes according to the assignment in the previous step, namely:
Figure FDA0002437046300000085
in the formula:
Figure FDA0002437046300000086
the final index weight coefficient is represented by,
Figure FDA0002437046300000087
obtaining a weight coefficient vector p of indexes in a final electric power spot market evaluation system 3Expressed as:
Figure FDA0002437046300000088
in the formula: p is a radical of3And the weight coefficient vector obtained by adopting a third weighting method, namely an entropy method based on an analytic hierarchy process is shown.
6. The comprehensive evaluation method of the electric power spot market according to claim 5, characterized in that: in the step S5, a comprehensive weight assignment method based on the game theory is used, and advantages of various weighting methods are combined to determine a comprehensive weight coefficient of each index, and the specific implementation method is as follows:
step 1: calculating a target weight vector ωres:
Figure FDA0002437046300000089
In the formula: omegaiRepresenting weight vectors obtained under the method in the ith under the k weight coefficient determination methods; alpha is alphaiThe linear combination coefficient represents the influence degree of the weight vector obtained under the method in the ith in the target weight vector;
step 2: representing target weight vector set as vector set in a centralized manner
Figure FDA0002437046300000091
And step 3: finding the optimal alphaiThe deviation between the set of target weight vectors and the respective weight vectors is minimized, i.e.:
Figure FDA0002437046300000092
the conditions for the optimized first derivative according to its differential properties are:
Figure FDA0002437046300000093
and 4, step 4: from the solved optimal alphaiRepresents the final integrated weight vector ωop。
7. The comprehensive evaluation method of the electric power spot market according to claim 6, characterized in that: s6 combines the gray characteristic of the electric power spot market, and uses a gray correlation degree analysis method to comprehensively evaluate the market, and the specific implementation method is as follows:
Step 1: standardizing each index data, processing according to a formula to obtain a result, and expressing each group of index data as the following power market operation effect evaluation matrix
Figure FDA0002437046300000094
In the formula: ma0Represents the evaluation vector of the electric power market operation effect in the theoretical operation state, st0jThe j-th index value under the theoretical operation state is generally regarded as 1, Ma by defaultiRepresents the ith electric power market operation effect evaluation vector, stijA value representing a j-th index in an i-th power market;
step 2: the evaluation matrix of the operation effect of the electric power market is obtained by multiplying the comprehensive index weight coefficient obtained by the game aggregation model, namely the evaluation matrix of the comprehensive operation effect of the electric power market is obtained
Figure FDA0002437046300000101
In the formula: CMa0Represents the comprehensive operation effect evaluation vector, Cst, of the power market in the theoretical operation state0jRepresents the j-th index comprehensive evaluation value, CMa, in the theoretical operating stateiRepresents the comprehensive operation effect evaluation vector, Cst, of the ith power marketijIndicates the j-th row index comprehensive evaluation value, omega, in the ith power marketjA comprehensive weight coefficient representing the j index;
and step 3: matrix for evaluating operation effect of power market by calculating difference array
Figure FDA0002437046300000102
Wherein
ΔCstij＝|Cst0j-Cstij|(i＝1,2,…,m；j＝1,2,…,n) (33)
In the formula: delta CMaiThe difference vector represents the comprehensive operation effect evaluation vector of the ith power market and the operation effect evaluation vector in the theoretical operation state;
And 4, step 4: solving the grey correlation coefficient of each index in each market and the corresponding index in the theoretical running state
Figure FDA0002437046300000111
In the formula, the value of delta m is the minimum value in the array matrix of the difference of the power market operation effect evaluation matrix
Figure FDA0002437046300000112
Delta M is the maximum value in the difference array matrix of the power market operation effect evaluation matrix, namely
Figure FDA0002437046300000113
Rho is a resolution coefficient and has a value range of [0, 1%]Influence the Gray correlation coefficient xiijBut has no influence on the final correlation coefficient ordering, and generally takes an empirical value of 0.618;
and 5: calculating the association degree of each market and the market under the theoretical operation state
Figure FDA0002437046300000114
Step 6: the obtained correlation degree gammaiSorting from big to small to obtain the sorting from good to bad about the operation condition of the i power markets and obtain a relative score value, namely a numerical value gamma of the relevancei。
CN202010255248.8A 2020-04-02 2020-04-02 Comprehensive evaluation method for electric power spot market Pending CN111861089A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202010255248.8A CN111861089A (en) 2020-04-02 2020-04-02 Comprehensive evaluation method for electric power spot market

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202010255248.8A CN111861089A (en) 2020-04-02 2020-04-02 Comprehensive evaluation method for electric power spot market

Publications (1)

Publication Number Publication Date
CN111861089A true CN111861089A (en) 2020-10-30

Family

ID=72985886

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202010255248.8A Pending CN111861089A (en) 2020-04-02 2020-04-02 Comprehensive evaluation method for electric power spot market

Country Status (1)

Country Link
CN (1) CN111861089A (en)

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112308427A (en) * 2020-11-02 2021-02-02 江苏省电力试验研究院有限公司 New energy consumption restriction factor evaluation method and system based on combined empowerment-grey correlation
CN112734080A (en) * 2020-12-09 2021-04-30 国网浙江省电力有限公司经济技术研究院 Power market operation efficiency evaluation method taking power grid as visual angle
CN112907955A (en) * 2021-01-28 2021-06-04 天津职业技术师范大学(中国职业培训指导教师进修中心) Evaluation method of vehicle-road cooperative system based on information fusion
CN114021905A (en) * 2021-10-19 2022-02-08 天元大数据信用管理有限公司 Credit risk evaluation method for small and medium-sized enterprises
CN115577902A (en) * 2022-09-02 2023-01-06 囯电南瑞南京控制系统有限公司 Inter-provincial electric power spot market evaluation index system construction method and system

Cited By (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112308427A (en) * 2020-11-02 2021-02-02 江苏省电力试验研究院有限公司 New energy consumption restriction factor evaluation method and system based on combined empowerment-grey correlation
CN112308427B (en) * 2020-11-02 2024-03-19 江苏省电力试验研究院有限公司 New energy consumption constraint factor evaluation method and system based on combined weighting-ash correlation
CN112734080A (en) * 2020-12-09 2021-04-30 国网浙江省电力有限公司经济技术研究院 Power market operation efficiency evaluation method taking power grid as visual angle
CN112907955A (en) * 2021-01-28 2021-06-04 天津职业技术师范大学(中国职业培训指导教师进修中心) Evaluation method of vehicle-road cooperative system based on information fusion
CN114021905A (en) * 2021-10-19 2022-02-08 天元大数据信用管理有限公司 Credit risk evaluation method for small and medium-sized enterprises
CN115577902A (en) * 2022-09-02 2023-01-06 囯电南瑞南京控制系统有限公司 Inter-provincial electric power spot market evaluation index system construction method and system
CN115577902B (en) * 2022-09-02 2023-09-22 囯电南瑞南京控制系统有限公司 Construction method and system of inter-provincial power spot market evaluation index system

Similar Documents

Publication Publication Date Title
CN111861089A (en) Comprehensive evaluation method for electric power spot market
CN109829604A (en) A kind of grid side energy-accumulating power station operational effect comprehensive estimation method
CN110689240A (en) Fuzzy comprehensive evaluation method for economic operation of power distribution network
CN110738435A (en) distribution network project investment decision evaluation method
CN107292534A (en) The yardstick competition evaluation method and device of urban power distribution network long term dynamics investment
CN108921376B (en) Optimal selection method and system for electricity reliability improvement object of intelligent power distribution network
CN106845777A (en) Case workload evaluation method, judge&#39;s performance evaluation techniques, modeling method, computing device and computer-readable recording medium
CN111798057B (en) Charging station site selection method based on fuzzy-hierarchy profit analysis
CN106056307B (en) Comprehensive benefit evaluation method for electric energy substitution technology based on comprehensive weighting method
CN108876114A (en) Consider the power grid efficiency benefit appraisal procedure of new energy access
CN113869616A (en) Self-adaptive residential user load prediction method
CN112465286A (en) Method for evaluating service capability of electric vehicle charging network
CN112488565A (en) Oversea engineering project risk evaluation method based on matter element extension
CN112085321A (en) Station area state evaluation method based on edge calculation
CN115759389A (en) Day-ahead photovoltaic power prediction method based on weather type similar day combination strategy
CN113327047B (en) Power marketing service channel decision method and system based on fuzzy comprehensive model
CN112330030B (en) System and method for predicting requirements of expansion materials
CN111563652A (en) Method for evaluating power grid enterprise comprehensive plan execution condition
CN104318488A (en) Method for pricing and compensating wind power AGC auxiliary services
CN110991715A (en) Power distribution network energy efficiency evaluation system and method based on improved grey correlation degree
CN115983664A (en) Comprehensive evaluation method for trading effect of energy storage participation in electric power market based on fuzzy analysis
Cheng et al. Deep convolutional neural network model for short-term electricity price forecasting
CN113033886B (en) Power distribution network planning construction evaluation method
CN114626714A (en) Power grid intelligent level evaluation method based on 4G/5G communication technology
CN114971272A (en) Medium-voltage distribution network gridding planning evaluation method and system

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
WD01 Invention patent application deemed withdrawn after publication

Application publication date: 20201030

WD01 Invention patent application deemed withdrawn after publication