Hybrid decision model-based power grid investment optimization method, electronic equipment and medium

Abstract

The invention discloses a power grid investment optimization method based on a hybrid decision model, electronic equipment and a medium, and the method comprises the steps: obtaining at least three power grid investment benefit evaluation indexes and at least three alternative investment projects; acquiring facial expressions of experts in the process of evaluating the power grid investment benefit evaluation indexes and the alternative investment projects; calculating the overall value of each alternative investment project based on a quantum spherical fuzzy M-SWARA algorithm, a quantum spherical fuzzy-ELCTRE algorithm and facial expressions; and sorting all the standby investment projects based on the overall value of the alternative investment projects to obtain a power grid investment optimization result. According to the method, the facial expression information of the decision-making experts is fused into the decision-making process, so that the credibility of expert judgment and the accuracy and reliability of investment optimization decision-making are improved, and accurate judgment of the priority of the power grid investment scheme can be realized, thereby assisting in making a more effective power grid investment strategy.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of power grid investment decision, and particularly relates to a power grid investment optimization method based on a hybrid decision model, an electronic device and a medium. BACKGROUND

[0002] In view of the characteristics of high capital intensity, strong system correlation and prominent risk uncertainty of power grid project investment, power enterprises urgently need to build a scientific power grid investment decision system to realize the continuous optimization of investment structure, the systematic improvement of capital use efficiency and the maximization of whole life cycle investment benefit, so as to effectively cope with the complex decision challenges in the construction of new power systems. Traditional power grid investment decision methods, such as cost-benefit analysis and single index evaluation, often fail to comprehensively consider technical, economic, social and environmental factors, and cannot effectively handle the complex interaction between factors and the fuzziness and subjectivity in the decision-making process.

[0003] In related technologies, multi-criteria decision methods are introduced into the field of power grid investment, such as the analytic hierarchy process, the network analysis method, the decision experiment and evaluation laboratory method, and the outranking distance method. For example, some establish a multi-level index system and a fuzzy comprehensive evaluation vector to evaluate the comprehensive effectiveness of a project. These methods to some extent realize the unification of qualitative and quantitative factors, and can better handle multi-objective and multi-criteria decision problems. However, existing power grid investment decision models mostly focus on the analysis of objective data and the structured expression of expert knowledge, pay less attention to the psychological state and behavior characteristics of decision-makers in the judgment process, do not consider the dynamic evaluation of expert judgment credibility in the decision-making process, and still rely completely on expert subjective weighting, making it difficult to ensure the objectivity and reliability of the evaluation results. SUMMARY

[0004] The purpose of the present application is to consider the dynamic evaluation of expert judgment credibility to more comprehensively participate in evaluation and improve the objectivity and reliability of the evaluation results.

[0005] To achieve the above purpose, the present application provides a power grid investment optimization method based on a hybrid decision model, comprising: obtaining power grid investment benefit evaluation indexes and alternative investment projects, the power grid investment benefit evaluation indexes and the alternative investment projects being at least three respectively; collecting facial expressions of experts in the process of evaluating the power grid investment benefit evaluation indexes and the alternative investment projects; calculating the overall value of each alternative investment project based on a quantum spherical fuzzy M-SWARA algorithm, a quantum spherical fuzzy-ELECTRE algorithm and the facial expressions; and sorting all the alternative investment projects based on the overall value of the alternative investment projects to obtain a power grid investment optimization result.

[0006] In an optional implementation, the overall value of each alternative investment project is calculated based on the quantum spherical fuzzy M-SWARA algorithm, the quantum spherical fuzzy ELECTRE algorithm and the facial expression, and specifically includes: constructing a facial expression correlation degree matrix based on the facial expression, the facial expression correlation degree matrix including a facial expression correlation degree matrix between multiple power grid investment benefit evaluation indexes and a facial expression correlation degree matrix between the power grid investment benefit evaluation indexes and the alternative investment projects; converting the facial expression correlation degree matrix into an index weight matrix based on the quantum spherical fuzzy M-SWARA algorithm; converting the facial expression correlation degree matrix into a weighted decision matrix based on the quantum spherical fuzzy ELECTRE algorithm and the index weight matrix; and calculating the overall value of each alternative investment project based on the weighted decision matrix.

[0007] In an optional implementation, the facial expression correlation degree matrix is constructed based on the facial expression, and specifically includes: identifying an emotion category based on the facial expression, the emotion category including contempt, surprise, surprise, an intermediate emotion between contempt and surprise, and an intermediate emotion between surprise and surprise; and constructing the facial expression correlation degree matrix based on the facial action coding and the emotion category.

[0008] In an optional implementation, the facial expression correlation degree matrix is converted into an index weight matrix based on the quantum spherical fuzzy M-SWARA algorithm, and specifically includes: converting the facial expression correlation degree matrix into a three-element quantum spherical fuzzy set based on the quantum spherical fuzzy M-SWARA algorithm to obtain a first fuzzy set matrix; aggregating the first fuzzy set matrix of multiple experts based on a spherical fuzzy algorithm to calculate an aggregated fuzzy set; performing defuzzification processing on the aggregated fuzzy set to obtain a defuzzification matrix; and performing normalization processing on the defuzzification matrix to obtain the index weight matrix.

[0009] In an optional implementation, the facial expression correlation degree matrix is converted into a three-element quantum spherical fuzzy set based on the spherical fuzzy algorithm and the quantum theory to obtain a first fuzzy set matrix, and specifically includes: calculating an element characteristic parameter of a facial action in the facial expression correlation degree matrix based on the spherical fuzzy algorithm, the element characteristic parameter including a membership degree, a non-membership degree and a hesitancy degree; calculating a phase angle corresponding to a degree value based on the quantum theory and the element characteristic parameter, the degree value being a specific numerical value of the element characteristic parameter; and obtaining the first fuzzy set matrix based on a combination of the facial action, the membership degree and the phase angle.

[0010] In an optional embodiment, the defuzzification matrix is ​​normalized to obtain the indicator weight matrix, which specifically includes: normalizing the defuzzification matrix to obtain an importance matrix between all the power grid investment benefit evaluation indicators and obtaining an importance value; based on the importance value, successively calculating the weight coefficients, adjustment weights and final weights between all the power grid investment benefit evaluation indicators; constructing an evaluation indicator relationship matrix between all the power grid investment benefit evaluation indicators based on the final weight; transposing the evaluation indicator relationship matrix and performing a power operation to obtain a stable matrix, the diagonal elements of the stable matrix being the weights corresponding to the power grid investment benefit evaluation indicators; and obtaining the indicator weight matrix based on the stable matrix.

[0011] In an optional embodiment, based on the quantum spherical fuzzy-ELECTRE algorithm and the indicator weight matrix, the facial expression correlation matrix is ​​converted into a weighted decision matrix, specifically including: based on the quantum spherical fuzzy-ELECTRE algorithm, the facial expression correlation matrix is ​​converted into a quantum spherical fuzzy set to obtain a second fuzzy set matrix; aggregating the fuzzy set matrices of each expert, and then performing defuzzification processing to convert the fuzzy set into a decision matrix, aggregating the second fuzzy set matrices of multiple experts; performing defuzzification processing on the second fuzzy set matrix to convert the quantum spherical fuzzy set into a decision matrix; and normalizing the decision matrix to obtain the weighted decision matrix.

[0012] In an optional embodiment, the overall value of each alternative investment project is calculated based on the weighted decision matrix, specifically including: based on the weighted decision matrix, constructing a consistency matrix and an inconsistency matrix according to a set of pros and cons classification indicators between any two of the multiple alternative investment projects; based on the consistency matrix, calculating the net superiority value of each of the alternative investment projects; based on the inconsistency matrix, calculating the net inferiority value of each of the alternative investment projects; taking the difference between the net superiority value and the net inferiority value to obtain a comprehensive ranking value, and obtaining the overall value.

[0013] The present invention also provides an electronic device comprising: at least one processor; a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute any one of the grid investment optimization methods based on the hybrid decision model.

[0014] The present invention also provides a medium storing a computer program, wherein when the computer program is executed by a processor, the method for optimizing power grid investment based on any one of the hybrid decision models is implemented.

[0015] The beneficial effects of the present application are: the present application introduces a mixed decision model of facial expression, integrates the facial expression information of decision experts into the decision process, improves the credibility of expert evaluation, and the accuracy, reliability and realization of investment optimization decision, and accurately judges the priority of power grid investment scheme, thereby assisting in formulating more effective power grid investment strategy. BRIEF DESCRIPTION OF DRAWINGS

[0016] Figure 1 A flowchart of a power grid investment optimization method based on a mixed decision model is provided for the implementation of the present application.

[0017] Figure 2 A block diagram of an electronic device is provided for the embodiments of the present application.

[0018] Explanation of reference numerals: 110, processor; 120, memory. DETAILED DESCRIPTION

[0019] When experts make complex judgments, psychological factors such as confidence level, hesitation degree, and cognitive load can significantly affect the reliability and consistency of their judgments. For example, experts may show hesitation when judging criteria with high uncertainty, or surprise when facing unexpected relationships. These emotional reactions contain deep information that traditional questionnaires or interviews cannot capture. Ignoring these psychological factors may lead to biased criterion weights, affecting the accuracy and robustness of the final decision.

[0020] The present application is further described in detail below through the accompanying drawings and specific embodiments.

[0021] As shown in the drawings, Figure 1 According to the embodiments of the present application, on the one hand, a power grid investment optimization method based on a mixed decision model is provided, comprising the following steps:

[0022] Step S101: obtaining power grid investment benefit evaluation indexes and alternative investment projects, wherein the power grid investment benefit evaluation indexes and the alternative investment projects are at least three respectively.

[0023] Step S103: collecting the facial expressions of experts during the evaluation of the power grid investment benefit evaluation indexes and the alternative investment projects.

[0024] Step S105: calculating the overall value of each alternative investment project based on quantum spherical fuzzy M-SWARA algorithm, quantum spherical fuzzy-ELECTRE algorithm and facial expression.

[0025] Step S107: sorting all alternative investment projects based on the overall value of the alternative investment projects to obtain the power grid investment optimization result.

[0026] In this embodiment, the power grid investment benefit evaluation indexes are at least three of safety and reliability, green and low carbon, digital intelligence, social responsibility and economic benefit, which can be referred to as indexes. Each candidate investment project needs to be evaluated from the above indexes, which can be referred to as a project. When experts evaluate the power grid investment benefit evaluation indexes and the candidate investment projects, the facial expression data of the experts is collected in real time. Facial expression recognition algorithms, such as a model based on a convolutional neural network, can be used to extract facial expression features of the experts.

[0027] According to the quantum spherical fuzzy M-SWARA algorithm, the mutual relationship between the indexes and the facial expression data of the experts can be comprehensively considered, and by using the quantum spherical fuzzy ELECTRE algorithm, the comprehensive performance of the candidate investment projects on the indexes can be more comprehensively reflected. For each candidate investment project, the overall value of each candidate investment project is calculated based on the performance of the candidate investment project on each evaluation index.

[0028] By collecting the facial expressions of the experts during the evaluation process, the subjectivity and flexibility of the decision-making process are increased, the decision-making result is closer to the actual demand, and in addition, the quantum spherical fuzzy theory, the M-SWARA algorithm and the ELECTRE algorithm are combined, the uncertainty and complexity in the decision-making process can be more comprehensively processed, all the candidate investment projects are sorted according to the calculated overall value, and the sorting result directly reflects the priority of each project. The calculation of the overall value comprehensively considers the index performance of the project, and can more accurately reflect the comprehensive benefit of the project. Based on the above sorting result, the optimization scheme of the power grid investment can be determined, and the benefits and risks of each investment project can be effectively balanced.

[0029] Further, in step S105, the overall value of each candidate investment project is calculated based on the quantum spherical fuzzy M-SWARA algorithm, the quantum spherical fuzzy-ELECTRE algorithm and the facial expression, and specifically includes the following steps:

[0030] In step S1051, a facial expression correlation degree matrix is constructed based on the facial expression, and the facial expression correlation degree matrix includes a facial expression correlation degree matrix between the multiple power grid investment benefit evaluation indexes and a facial expression correlation degree matrix between the power grid investment benefit evaluation indexes and the candidate investment projects.

[0031] In step S1053, the facial expression correlation degree matrix is converted into an index weight matrix based on the quantum spherical fuzzy M-SWARA algorithm.

[0032] In step S1055, the facial expression correlation degree matrix is converted into a weighted decision matrix based on the quantum spherical fuzzy-ELECTRE algorithm and the index weight matrix.

[0033] In step S1057, the overall value of each candidate investment project is calculated based on the weighted decision matrix.

[0034] In this embodiment, the facial expression correlation matrix between power grid investment benefit evaluation indexes is used to measure the correlation between different evaluation indexes and is quantified through the facial expressions of experts in the evaluation process. The facial expression correlation matrix between power grid investment benefit evaluation indexes and alternative investment projects is used to measure the correlation between each evaluation index and alternative investment projects.

[0035] The element values of the correlation matrix are converted from facial expression data, for example, by calculating the similarity or correlation of expert expressions to determine the correlation. The quantum spherical fuzzy ELECTRE algorithm combines quantum spherical fuzzy theory and the ELECTRE algorithm to further convert the index weight matrix into a weighted decision matrix. In the weighted decision matrix, the performance of each alternative investment project on each evaluation index is assigned a corresponding weight, which reflects the importance of the index and the facial expression data of the expert. Based on the weighted decision matrix, the overall value of each alternative investment project can be calculated, which combines the facial expression data of the expert and the quantum spherical fuzzy algorithm and reflects the comprehensive performance of the alternative investment project on all evaluation indexes.

[0036] Further, in step S1051, a facial expression correlation matrix is constructed based on facial expressions, specifically including the following steps:

[0037] In step S10511, an emotion category is identified based on facial expressions, and the emotion category includes contempt, surprise, surprise, intermediate emotion between contempt and surprise, and intermediate emotion between surprise and surprise.

[0038] In step S10512, a facial expression correlation matrix is constructed based on facial action coding and emotion categories.

[0039] In the process of evaluating the investment benefit of the power grid, a number of experts are invited to form a decision maker set to judge the mutual relationship between the indexes and the influence of the investment projects on the indexes. The facial expressions of the experts can reflect their subjective feelings about different indexes and projects. In the emotion categories defined in this method, contempt indicates contempt or disapproval of a certain index or project, surprise indicates surprise or attention to a certain index or project, surprise indicates high recognition and satisfaction for a certain index or project, intermediate emotion between contempt and surprise indicates the complex emotion of the expert towards a certain index or project, and intermediate emotion between surprise and surprise indicates the complex emotion of the expert towards a certain index or project.

[0040] The mapping relationship between the emotion category, facial expression, and action unit is as follows:

[0041]

[0042] The Facial Action Coding System (FACS) is a method for describing facial muscle movements. Through FACS, an expert's facial expression can be broken down into a series of basic action units (AUs), and the intensity and duration of each action unit is recorded.

[0043] The facial expressions of the decision-making experts during the evaluation process are captured, the facial action coding system is applied to analyze their non-verbal behaviors, the facial expressions at each time are translated into action units expressed in mathematics, so that the emotions of the experts are quantified, and the facial expression correlation matrix between the indicators and the facial expression correlation matrix between the indicators and the investment projects are constructed based on the two-by-two combination of the action units.

[0044] Among them, for each pair of power grid investment benefit evaluation indicators, the facial expression actions of the experts when evaluating the two indicators are recorded. According to the most obvious two actions and their corresponding emotion categories, the correlation between the two indicators is calculated. For example, if the experts show a happy emotion when evaluating the two indicators, the correlation between the two indicators is higher.

[0045] For each power grid investment benefit evaluation indicator and each alternative investment project, the facial expression actions of the experts when evaluating the indicator and the project are recorded. Similarly, the most obvious two actions and their corresponding emotion categories are calculated to calculate the impact of the indicator on the project.

[0046] The emotion categories and facial action coding are converted into numerical correlation. The emotion categories can be mapped to numerical action unit combinations, such as [5, 6], and finally a matrix with each action unit combination as an element is formed.

[0047] Through the above steps, based on facial expression recognition and emotion categories, a facial expression correlation matrix reflecting the subjective evaluation of the experts can be constructed.

[0048] Further, in step S1053, the facial expression correlation matrix is converted into an indicator weight matrix based on the quantum spherical fuzzy M-SWARA algorithm, which includes the following steps:

[0049] Step S10531: Based on the quantum spherical fuzzy M-SWARA algorithm, the facial expression correlation matrix is converted into a three-tuple quantum spherical fuzzy set to obtain a first fuzzy set matrix.

[0050] Step S10533: Based on the spherical fuzzy algorithm, the first fuzzy set matrices of multiple experts are aggregated to calculate an aggregated fuzzy set.

[0051] Step S10535: The aggregated fuzzy set is de-fuzzified to obtain a de-fuzzified matrix.

[0052] Step S10537: Normalizing the defuzzification matrix to obtain the index weight matrix.

[0053] In this embodiment, the first fuzzy set matrix is calculated based on the quantum spherical fuzzy M-SWARA algorithm, which can quantify the uncertainty. Assuming that multiple experts evaluate the same set of power grid investment benefit evaluation indicators and alternative investment projects based on facial expressions, each expert will generate a first fuzzy set matrix. Using the spherical fuzzy algorithm, the first fuzzy set matrices of all experts are aggregated. During the aggregation process, the similarity and difference between the fuzzy sets of each expert are considered, and a comprehensive aggregated fuzzy set is obtained through weighted averaging or other aggregation strategies. Aggregating the fuzzy set matrices of multiple experts can integrate the opinions of different experts and reduce the deviation of single expert opinions, thereby obtaining a more representative and consistent fuzzy set and improving the consistency of information.

[0054] Based on steps S10533 and S10535, the fuzzy set matrices of each expert are aggregated following the spherical fuzzy operation criteria, and then defuzzification is performed to convert the fuzzy set into a specific numerical expression that can be sorted. The purpose of defuzzification is to convert the uncertainty information in the fuzzy set into explicit numerical values to facilitate subsequent calculation and analysis.

[0055] wherein the fuzzy set is aggregated based on the following formula:

[0056]

[0057] Defuzzification is performed based on the following formula:

[0058]

[0059] wherein s is the aggregated fuzzy set, is the triple component of the quantum spherical fuzzy set of the i-th expert, a i , β i , γ i is the corresponding phase angle of the triple component.

[0060] Each element in the defuzzification matrix is normalized to fall within a unified range, which can eliminate the dimensional and order-of-magnitude differences between different indicators.

[0061] Further, in step S10531, the facial expression correlation degree matrix is converted into a triple quantum spherical fuzzy set based on the spherical fuzzy algorithm and quantum theory to obtain the first fuzzy set matrix, which specifically includes the following steps:

[0062] Step S105311: Based on the spherical fuzzy algorithm, the element characteristic parameters of the facial expression correlation degree matrix are calculated, including membership, non-membership and hesitation.

[0063] Step S105313: Based on the quantum theory and the element characteristic parameters, the phase angle corresponding to the degree value is calculated, and the degree value is the specific numerical value of the element characteristic parameter.

[0064] Step S105315: Based on the combination of facial actions, membership and phase angle, the first fuzzy set matrix is obtained.

[0065] Based on step S10531, first, for each element in the facial expression correlation degree matrix, such as the correlation degree between the action unit combination or the emotion category, it is represented as a three-tuple form, i.e. membership, non-membership and hesitation, by using quantum spherical fuzzy theory. The phase angle corresponding to each degree value is calculated by quantum theory, each correlation degree element is converted into the form of quantum spherical fuzzy set, and a matrix containing quantum spherical fuzzy set, i.e. the first fuzzy set matrix, is generated. Through the three-tuple form of membership, non-membership and hesitation, the uncertainty in the expert facial expression data can be more comprehensively quantified.

[0066] Among them, the calculated membership corresponding to each action unit combination and the quantum spherical fuzzy set are as follows:

[0067]

[0068] Further, based on step S10537, the defuzzification matrix is normalized to obtain the index weight matrix, which includes the following steps:

[0069] Step S105371: The defuzzification matrix is normalized to obtain the importance matrix between all power grid investment benefit evaluation indexes, and the importance value is obtained.

[0070] Step S105373: Based on the importance value, the weight coefficients, adjusted weights and final weights between all power grid investment benefit evaluation indexes are calculated in sequence.

[0071] Step S105375: Based on the final weight, the evaluation index relationship matrix between all power grid investment benefit evaluation indexes is constructed.

[0072] Step S105377: The evaluation index relationship matrix is transposed and operated by power to obtain the stability matrix, and the diagonal elements of the stability matrix are the weights corresponding to the power grid investment benefit evaluation indexes.

[0073] Step S105379: Based on the stability matrix, the index weight matrix is obtained.

[0074] In this embodiment, the normalized disambiguation matrix is used to calculate the weight coefficient of each evaluation index. The weight coefficient reflects the relative importance of each index in the decision-making process. The weight calculation method based on quantum spherical fuzzy set can be used to determine the weight coefficient by calculating the relative importance value between each index. Through matrix transposition and power operation, more stable weight values can be obtained, avoiding the decision bias caused by the instability of the matrix. The index weight matrix integrates the facial expression data of experts and quantum spherical fuzzy theory, and can provide comprehensive decision support for power grid investment optimization.

[0075] wherein, based on the importance matrix, the weight coefficients k between each index are calculated in turn j , the adjustment weight q j and the maximum weight w j , and the relationship matrix between each index and the impact direction are constructed from the maximum weight.

[0076] wherein, the weight coefficient, the adjustment weight and the maximum weight are calculated based on the following formula:

[0077]

[0078] In the formula, s j is the importance value.

[0079] By calculating the weight coefficient, the importance of each evaluation index in the decision-making process can be more scientifically reflected. The calculation of the weight coefficient combines the subjective evaluation of experts and objective data, and further quantifies the uncertainty in the decision-making process.

[0080] In step S1055, based on the quantum spherical fuzzy-ELECTRE algorithm and the index weight matrix, the facial expression correlation degree matrix is converted into a weighted decision matrix, which specifically includes the following steps:

[0081] In step S10551, based on the quantum spherical fuzzy-ELECTRE algorithm, the facial expression correlation degree matrix is converted into a quantum spherical fuzzy set to obtain a second fuzzy set matrix.

[0082] In step S10553, the fuzzy set matrices of each expert are aggregated, and then disambiguation processing is performed to convert the fuzzy set into a decision matrix, and the second fuzzy set matrices of multiple experts are aggregated.

[0083] In step S10555, the second fuzzy set matrix is disambiguated to convert the quantum spherical fuzzy set into a decision matrix.

[0084] In step S10557, the decision matrix is normalized to obtain a weighted decision matrix.

[0085] In this embodiment, each quantum spherical fuzzy set element in the second fuzzy set matrix is ​​defuzzified. Through defuzzification, information such as membership, non-membership, and hesitation in the fuzzy set is converted into specific numerical values, forming a clear decision matrix. Defuzzification makes the information in the decision matrix clearer and more specific, facilitating subsequent weight calculation and ranking. After converting fuzzy information into numerical values, the consistency of the data in the decision matrix is ​​ensured.

[0086] The normalized decision matrix is ​​weighted using the Quantum Spherical Fuzzy-ELECTRE algorithm, combined with the expert weighting of each indicator. The weight of each indicator reflects its relative importance in the decision-making process. This weighting process results in the final weighted decision matrix. This weighting combines the expert's subjective judgment with objective data, making the decision matrix more scientifically reflect the relationship between each indicator and project.

[0087] Among them, normalization is performed based on the following formula:

[0088]

[0089] Where, X ij are the values ​​of each element in the decision matrix.

[0090] Construct a weighted decision matrix based on the following formula:

[0091] v ij =w ij ×r ij ;

[0092] Where w ij are the values ​​of each element in the stable matrix W.

[0093] Step S1057, calculating the overall value of each investment candidate based on the weighted decision matrix, specifically includes the following steps:

[0094] Step S10571: Based on the weighted decision matrix, a consistency matrix and an inconsistency matrix are constructed according to a set of indicators for dividing the pros and cons between any two of the multiple investment alternatives.

[0095] Step S10573: Based on the consistency matrix, calculate the net merit for each alternative investment project.

[0096] Step S10575: Based on the inconsistency matrix, calculate the net disadvantage value for each alternative investment project.

[0097] Step S10577: Subtract the net superiority value from the net inferiority value to obtain a comprehensive ranking value and an overall value.

[0098] In this embodiment, the weighted decision matrix is analyzed to extract the weighted score of each alternative investment project on each evaluation index. For each alternative investment project, its performance in the weighted decision matrix is analyzed to determine the set of superior and inferior indexes compared with other projects. For each pair of alternative investment projects i and j, their scores on each evaluation index are compared to determine on which indexes project i is superior to project j and on which indexes project i is inferior to project j. For example, if project A is superior to project B on an index, the index belongs to the "superior index set" of project A; if project A is inferior to project B on an index, the index belongs to the "inferior index set" of project A.

[0099] For each pair of projects i and j, two sets are defined:

[0100] Superior index set A ij : contains all evaluation indexes on which project i is superior to project j.

[0101] Inferior index set B ij : contains all evaluation indexes on which project i is inferior to project j.

[0102] The consistency matrix is used to reflect the relative superiority relationship between alternative investment projects. For each pair of projects i and j, the consistency index value of project i relative to project j is calculated according to its superior index set. The element Cij of the consistency matrix represents the degree of superiority of project i relative to project j, which is usually obtained by expert scoring or calculation method.

[0103] The inconsistency matrix is used to reflect the relative inferiority relationship between alternative investment projects. Also for each pair of projects i and j, the inconsistency index value of project i relative to project j is calculated according to its inferior index set. The element D ij of the inconsistency matrix represents the degree of inferiority of project i relative to project j.

[0104] wherein the consistency matrix and the inconsistency matrix are constructed based on the following formula:

[0105]

[0106] c0 ab = {j | v aj > v bj} ;

[0107] d0 ab = {j | v aj < v bj} ;

[0108]

[0109] wherein C, D are the consistency and inconsistency matrices respectively, c ab , dab respectively are consistency and inconsistency indexes between investment projects.

[0110] For each investment project, the net superior value is calculated based on the consistency matrix, the net inferior value is calculated based on the inconsistency matrix, and the comprehensive ranking value is obtained by the difference, and the ranking of various types of investment projects is performed.

[0111] The net superior value and the net inferior value are calculated based on the following formula:

[0112]

[0113] In the formula, c a , d a are the net superior value and the net inferior value respectively.

[0114] The calculation of the net superior value quantifies the advantage degree of the project, which can provide an important index for the comprehensive evaluation of the project, and can intuitively compare the advantages of different projects. The calculation of the net inferior value quantifies the disadvantage degree of the project, which provides an important index for the comprehensive evaluation of the project. Through the net inferior value, the potential disadvantage of the project in some indexes can be identified, which helps the decision maker to better evaluate the risk.

[0115] The comprehensive ranking value Ea considers both the advantages and disadvantages of the project, and can more comprehensively reflect the overall value of the project.

[0116] For each alternative investment project i, the comprehensive ranking value Ea of the project is calculated, that is, the difference between the net superior value and the net inferior value:

[0117] Ea=c a -d a ;

[0118] Through the comprehensive ranking value, all alternative investment projects can be ranked, so as to determine the optimal investment project.

[0119] In another aspect, the present application also provides an electronic device, comprising: at least one processor 110; a memory 120 in communication connection with the at least one processor; wherein the memory 120 stores instructions executable by the at least one processor 110, and the instructions are executed by the at least one processor 110 to enable the at least one processor 110 to perform any one of the power grid investment optimization methods based on the hybrid decision model.

[0120] In another aspect, the present application also provides a computer storage medium, which stores a computer program, and the computer program is executed by a processor to implement any one of the power grid investment optimization methods based on the hybrid decision model.

[0121] The computer storage medium can be referred to as a medium simply. Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be completed by a computer program instructing related hardware. The program can be stored in a non-volatile computer readable storage medium, and when the program is executed, the processes of the above-mentioned embodiments can be included. Any reference to memory, storage, database or other medium used in each embodiment provided by the present application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration but not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data SDRAM (DDR SDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM). Each embodiment in the specification is described in a progressive manner, and the same or similar parts of each embodiment can be referred to each other. Each embodiment focuses on the difference from other embodiments. Especially, for the device, equipment, non-volatile computer storage medium embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant part can be referred to the part of the method embodiment.

[0122] The above embodiments are only examples for clearly illustrating, and not limiting the embodiments. Based on the above description, other different forms of changes or variations can be made by those skilled in the art. Here, all the embodiments cannot be exhausted, and the obvious changes or variations still fall within the protection scope of the present application.

Claims

1. A power grid investment optimization method based on a hybrid decision model, characterized in that: include: Obtaining power grid investment benefit evaluation indicators and alternative investment projects, wherein the power grid investment benefit evaluation indicators and the alternative investment projects are at least three; collecting facial expressions of experts during the process of evaluating the power grid investment benefit evaluation indicators and the alternative investment projects; Calculate the overall value of each investment candidate based on the quantum spherical fuzzy M-SWARA algorithm, the quantum spherical fuzzy-ELECTRE algorithm and the facial expression; All the alternative investment projects are ranked based on the overall value of the alternative investment projects to obtain a power grid investment optimization result.

2. The power grid investment optimization method based on the hybrid decision model according to claim 1 is characterized in that: The overall value of each investment candidate is calculated based on the quantum spherical fuzzy M-SWARA algorithm, the quantum spherical fuzzy-ELECTRE algorithm and the facial expression, specifically including: Constructing a facial expression correlation matrix based on the facial expressions, the facial expression correlation matrix including a facial expression correlation matrix between a plurality of power grid investment benefit evaluation indicators and a facial expression correlation matrix between the power grid investment benefit evaluation indicators and alternative investment projects; Based on the quantum spherical fuzzy M-SWARA algorithm, the facial expression correlation matrix is ​​converted into an indicator weight matrix; Based on the quantum spherical fuzzy-ELECTRE algorithm and the indicator weight matrix, the facial expression correlation matrix is ​​converted into a weighted decision matrix; The overall value of each investment alternative project is calculated based on the weighted decision matrix.

3. The power grid investment optimization method based on the hybrid decision model according to claim 2 is characterized in that: Constructing a facial expression correlation matrix based on the facial expressions specifically includes: identifying emotion categories based on the facial expressions, the emotion categories comprising contempt, surprise, delight, an intermediate emotion between contempt and surprise, and an intermediate emotion between delight and delight; The facial expression correlation matrix is ​​constructed based on facial action coding and the emotion category.

4. The power grid investment optimization method based on the hybrid decision model according to claim 2 is characterized in that: Based on the quantum spherical fuzzy M-SWARA algorithm, the facial expression correlation matrix is ​​converted into an indicator weight matrix, specifically including: Based on the quantum spherical fuzzy M-SWARA algorithm, the facial expression correlation matrix is ​​converted into a triple quantum spherical fuzzy set to obtain a first fuzzy set matrix; Aggregating the first fuzzy set matrices of multiple experts based on a spherical fuzzy algorithm to calculate an aggregated fuzzy set; Defuzzification is performed on the aggregated fuzzy set to obtain a defuzzification matrix; The defuzzification matrix is ​​normalized to obtain the indicator weight matrix.

5. The power grid investment optimization method based on the hybrid decision model according to claim 4 is characterized in that: Based on the spherical fuzzy algorithm and quantum theory, the facial expression correlation matrix is ​​converted into a triple quantum spherical fuzzy set to obtain a first fuzzy set matrix, which specifically includes: Calculating element feature parameters of facial movements in the facial expression correlation matrix based on a spherical fuzzy algorithm, wherein the element feature parameters include membership, non-membership, and hesitation; Calculating a phase angle corresponding to a degree value based on quantum theory and the characteristic parameter of the element, wherein the degree value is a specific numerical value of the characteristic parameter of the element; The first fuzzy set matrix is ​​obtained based on the combination of facial actions, the membership degree and the phase angle.

6. The power grid investment optimization method based on the hybrid decision model according to claim 4 is characterized in that: Normalizing the defuzzification matrix to obtain the indicator weight matrix specifically includes: Normalizing the defuzzification matrix to obtain an importance matrix between all the power grid investment benefit evaluation indicators and obtain an importance value; Based on the importance value, sequentially calculating weight coefficients, adjustment weights and final weights among all the power grid investment benefit evaluation indicators; Constructing an evaluation index relationship matrix among all the power grid investment benefit evaluation indexes based on the final weights; Transpose the evaluation index relationship matrix and perform a power operation to obtain a stability matrix, wherein the diagonal elements of the stability matrix are the weights corresponding to the power grid investment benefit evaluation index; The indicator weight matrix is ​​obtained based on the stability matrix.

7. The power grid investment optimization method based on the hybrid decision model according to claim 6 is characterized in that: Based on the quantum spherical fuzzy-ELECTRE algorithm and the indicator weight matrix, the facial expression correlation matrix is ​​converted into a weighted decision matrix, specifically including: Based on the quantum spherical fuzzy-ELECTRE algorithm, the facial expression correlation matrix is ​​converted into a quantum spherical fuzzy set to obtain a second fuzzy set matrix; Aggregating the fuzzy set matrices of each expert, performing defuzzification processing, converting the fuzzy set into a decision matrix, and aggregating the second fuzzy set matrices of multiple experts; performing defuzzification processing on the second fuzzy set matrix to convert the quantum spherical fuzzy set into a decision matrix; The decision matrix is ​​normalized to obtain the weighted decision matrix.

8. The power grid investment optimization method based on the hybrid decision model according to claim 7 is characterized in that: The overall value of each alternative investment project is calculated based on the weighted decision matrix, specifically including: Based on the weighted decision matrix, constructing a consistency matrix and an inconsistency matrix according to a set of pros and cons of any two of the plurality of alternative investment projects; Calculating a net merit for each of the alternative investment projects based on the consistency matrix; Calculating a net disadvantage value for each of the alternative investment projects based on the inconsistency matrix; The net superior value and the net inferior value are subtracted to obtain a comprehensive ranking value to obtain the overall value.

9. An electronic device, characterized in that: include: at least one processor; A memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor so as to enable the at least one processor to execute the power grid investment optimization method based on the hybrid decision model as described in any one of claims 1 to 8.

10. A medium, characterized in that A computer program is stored, and when the computer program is executed by a processor, a power grid investment optimization method based on a hybrid decision model as described in any one of claims 1 to 8 is implemented.

Images