Novel power system comprehensive evaluation method based on hierarchical evidence reasoning

Abstract

The invention provides a novel power system comprehensive evaluation method based on hierarchical evidence reasoning. The novel power system comprehensive evaluation method comprises the following steps of S1, selecting an index of an lth layer; s2, constructing an index pairwise comparison matrix C and an index data matrix X at the same time; s3, respectively calculating a subjective weight W1 and an objective weight W2 of the index by adopting an analytic hierarchy process and a CRITIC method; s4, calculating the comprehensive weight W of the index by using Bayesian network reasoning; s5, selecting index data of the kth region; s6, calculating a utility average value uavg (Il + 1k) of the (l + 1) th layer of indexes by adopting an evidence reasoning method; s7, judging whether the region is the last region or not; and S8, judging whether the index is the last layer index or not. By applying the technical scheme, subjective judgment and objective analysis can be integrated to realize multi-layer fusion of evidences in an uncertain environment, a novel electric power system statistical index system with a multi-layer structure can be accurately evaluated, and the comprehensive evaluation score of the novel electric power system in each region can be obtained.

Description

Technical Field

[0001] The invention relates to the technical field of comprehensive evaluation of electric power systems, in particular to a novel comprehensive evaluation method of electric power systems based on hierarchical evidence reasoning. Background Art

[0003] In the process of building new power systems, new business forms and new elements such as electric vehicles, new energy storage and smart grids are developing rapidly, and the structure and characteristics of traditional power systems are constantly changing. The rapidly developing new business forms and new elements lack corresponding statistical standards and norms. Some traditional power system statistical indicators are difficult to adapt to the rapidly changing power system, which greatly increases the ambiguity and diversity of the statistical indicator data collection process, making the comprehensive evaluation of new power systems uncertain. However, the traditional comprehensive evaluation method does not take into account uncertain information and ignores the impact of uncertain information on the evaluation, which makes the credibility and accuracy of the evaluation results low and cannot effectively serve the new power system. Summary of the invention

[0004] In view of this, the purpose of the present invention is to provide a new comprehensive evaluation method for power systems based on hierarchical evidential reasoning, which can realize the effective integration of certain and uncertain information, improve the scientificity and interpretability of the new comprehensive evaluation method for power systems, and accurately evaluate the overall and various aspects of the construction of new power systems in various regions.

[0005] To achieve the above object, the present invention adopts the following technical solution: a new comprehensive evaluation method of power system based on hierarchical evidence reasoning, comprising the following steps:

[0006] Step S1: Select the first level indicator I of the new power system statistical indicator system l ;

[0007] Step S2: Use Satty comparison scale and expert survey method to compare the first level indicator I l Construct a pairwise comparison matrix C; at the same time, construct an indicator data matrix X based on the indicator data of the first layer;

[0008] Step S3: Use the hierarchical analysis method to calculate and analyze the pairwise comparison matrix and output the subjective weight W 1 At the same time, the CRITIC weight method is used to calculate the indicator data matrix X and output the objective weight W 2 ;

[0009] Step S4: Use Bayesian network reasoning to derive the subjective weight W 1 and objective weight W 2 Perform combined weighting and output comprehensive weight W (n) ;

[0010] Step S5: selecting the index data of the first layer of the statistical index system of the new power system in the kth region;

[0011] Step S6: Use the evidential reasoning method to calculate the utility average u of the l+1th layer indicator avg (I l+1 k ), as the indicator data of the l+1th layer indicator;

[0012] Step S7: Determine whether the current region is the last region; if not, then k+1, execute steps S5 to S6; if yes, then execute the next step;

[0013] Step S8: Determine whether the current indicator level is the last level indicator; if not, then l+1, execute steps S1 to S7; if yes, then output the utility average value u corresponding to all regions avg (A) As a comprehensive evaluation score of new power systems in all regions.

[0014] In a preferred embodiment, the step S2 comprises the following steps:

[0015] Step S21: Establish a hierarchical structure based on the subordinate relationship of two adjacent layers of indicators: take the statistical indicator data of the new power system in each region as the scheme layer; take the indicator system as the criterion layer, assuming that the indicator system has l max Layer index, then the criterion layer is l max Layer, where all n indicators in the lth layer of the indicator system are y l ={y l (1) ,y l (2) ,...,y l (n)}; The construction of new power systems is taken as the target layer;

[0016] Step S22: Using the Satty comparison scale, invite researchers to compare the n indicators y in the first layer according to their own understanding. l (n) Score the importance of indicators;

[0017] Step S23: Compare all n indicators in the first layer in pairs and assign indicator y l ={y l (1) ,y l (2) ,...,y l (n) For the Jth criterion CR in the l+1th layer k The importance of taking two elements at a time and Use cij express and CR J The comparison results are used to construct a judgment matrix for pairwise comparison or pairwise comparison. Represents; where Z is the judgment matrix C J The order of is determined by the number of indicators in the lth layer belonging to the kth indicator in the l+1th layer;

[0018] Step S24: Obtain all n indicators y of the lth layer in all k regions l The indicator data of the first layer is obtained by querying statistical data, and the indicator data of the remaining layers are obtained from the indicators of the l-1 layer through the evidence reasoning method; construct the indicator data matrix X = (x ij ) n×k , where x ij (i∈(1,n),j∈(1,k)) represents the data value of indicator i in the new power system statistical indicator system I of the jth region; n represents the number of indicators; k represents the number of regions.

[0019] Step S25: normalize the indicator data matrix X to eliminate the dimension:

[0020] When the indicator is a positive indicator, the normalization method is:

[0021]

[0022] When the indicator is a negative indicator, the normalization method is:

[0023]

[0024] In the formula, x ij represents the data value of the i-th indicator in the j-th region; x ij ' represents the normalized data value of the ith indicator in the jth region; represents the minimum value in index i, represents the maximum value in index i;

[0025] Get the normalized indicator data matrix X′=(x i ' j ) n×k .

[0026] In a preferred embodiment, the step 3 comprises the following steps:

[0027] Step S31: Calculate the judgment matrix C J Find the largest eigenvalue λ among the eigenvalues max , calculate the consistency index C·I of the judgment matrix according to formula (3):

[0028]

[0029] Step S32: Query the average random consistency index R·I:

[0030] Step S33: Calculate the consistency ratio C·R according to formula (4):

[0031]

[0032] When C·R<0.1, it is considered that the judgment matrix C J The consistency is acceptable; otherwise, researchers need to re-conduct pairwise comparisons and make judgments on the matrix C. J Adjust it so that it can pass the consistency test;

[0033] Step S34: According to the judgment matrix C J and the maximum eigenvalue λ max From formula C J ω J,1 =λ max ω J,1 Calculating λ max The corresponding eigenvector ω J,1 , that is, the eigenvector is It is composed of the subjective weights of the Z indicators in the lth layer that belong to the Jth indicator in the l+1th layer;

[0034] Step S35: According to the normalized index data matrix X′=(x i ' j ) n×k The variability of the indicator data is calculated by formula (5):

[0035]

[0036] In the formula, represents the average value of index i, x′ ij represents the data value of indicator i in the jth region after normalization, S i represents the standard deviation of index i;

[0037] Step S36: Calculate the conflict of indicator data using formula (6):

[0038]

[0039] In the formula, R i represents the correlation of indicator i, r zi represents the correlation coefficient between indicator i and indicator z;

[0040] Step S37: Calculate the information volume of the index:

[0041]

[0042] In the formula, C i Indicates the amount of information contained in indicator i. The more information an indicator has, the more weight coefficients it is assigned. i Represents the variability of indicator i.

[0043] Step S38: Calculate the index weight:

[0044]

[0045] Repeat steps S35 to S38 in sequence to obtain the objective weights of the Z l-th layer indicators that belong to the J-th indicator in the l+1-th layer, and together form the objective weight vector

[0046] Step S39: Repeat steps S31 to S38 to obtain the subjective weights and objective weights of all indicators in the first layer of the new power system statistical indicator system.

[0047] In a preferred embodiment, step S4 comprises the following steps:

[0048] Step S41: Assume that the subjective weight W of the i-th indicator in the l-th layer of the new power system statistical indicator system is 1 (i) The corresponding subjective weight coefficient is l 1 (i) , objective weight W 2 (i) The corresponding objective weight coefficient is l 2 (i) ;

[0049] Step S42: constructing a Bayesian network structure, setting the subjective and objective weights of indicator i as the parent node, and setting the comprehensive weight as the child node;

[0050] Step S43: Set the prior probability distribution of each node in the Bayesian network, where the prior probability distribution of the first parent node is λ 1 (i) , the prior probability distribution of the second parent node is λ 2 (i) ,Using uniform distribution as the prior probability distribution of comprehensive weights, thus determining the complete Bayesian network;

[0051] Step S44: using the BNT toolbox, then determining the reasoning engine, updating the probability distribution of the parent node according to the process from effect to cause, that is, the value of the known child node, so as to perform reverse reasoning using the backward reasoning method;

[0052] Step S45: Convert the Bayesian network into a joint tree: Convert the Bayesian network into a secondary structure, i.e., a joint tree, and then perform probability calculation through a message passing process defined on the joint tree;

[0053] Step S46: using a tree-shaped joint probability inference algorithm to update the edge probability distribution of the node by passing messages between the joint tree nodes, wherein the evidence of the comprehensive weight is set to 1;

[0054] Step S47: Local update, and finally obtain the global reasoning result, that is, the posterior probability λ of the parent node is inferred 1 (i*) and λ 2 (i*) ;

[0055] Step S48: Calculate the comprehensive weight W of index i (i) :

[0056] W (i) =λ 1 (i*) W 1 (i) +λ 2 (i*) W 2 (i) (9)

[0057] Step S49: Repeat steps S41 to S48 to obtain the comprehensive weights of all indicators in the first layer of the new power system statistical indicator system.

[0058] In a preferred embodiment, step S6 comprises the following steps:

[0059] Step S61: Construct an evaluation set L with a total of h evaluation levels = {L 1 ,L 2 ,...,L h}, then the evaluation set of all n indicators in the kth region and the lth layer is y k,l ={y k,l (1) ,y k,l (2) ,...,y k,l (n)};

[0060] Step S62: Set the best and worst indicator evaluation matrices respectively and According to formula (10), the indicator evaluation matrix of the indicator i at the lth level of the kth region is obtained based on the indicator data.

[0061]

[0062] Step S63: For the index i of the kth region and the lth level, its index evaluation matrix Combined with the corresponding index weight W (i) Multiply them together to get the basic credibility matrix of the indicator Right now

[0063] Step S64: Combine the weights W (i) And the index evaluation vector According to equations (11) to (13), the uncertainty basic credibility matrix of indicator i is obtained:

[0064]

[0065] In the formula, represents the residual uncertainty of the ith indicator in the kth region; It represents the uncertainty after weighted allocation of the ith indicator in the kth region; It represents the confidence of the i-th indicator in the k-th region at the j-th evaluation level.

[0066] Step S65: According to equations (14) to (18), the Z index evaluation matrices of the kth region belonging to the index J in the l+1th layer are merged to calculate the total basic credibility matrix of the index J: And the total uncertainty basic credibility matrix

[0067]

[0068] In the formula, I(i) represents the comprehensive index after integrating the first i indicators. is the total basic credibility of the unassigned indicator I(i), that is, the total uncertainty basic credibility assigned to indicator I(i), I(i+1) represents the comprehensive indicator after integrating the first i+1 indicators, is the total basic credibility of indicator I(i+1) at evaluation level t;

[0069] Step S66: Fuse the Z index evaluation matrix of the first layer to obtain the total basic credibility matrix of the index J of the first layer + 1 and the total uncertainty basic credibility According to formula (19) and formula (20), the total confidence of the l+1th layer indicator J at each evaluation level is calculated as and the total uncertainty confidence

[0070]

[0071] Step S67: Determine the evaluation level set L = {L 1 ,L 2 ,...,L h The corresponding utility value set is u(L)={u(L 1 ),u(L 2 ),...,u(L h )}, according to formula (21) to formula (23), the utility average value of the l+1th level indicator J in the kth region can be calculated

[0072]

[0073] Step S68: Repeat steps S51 to S57 to calculate the utility average of all indicators in the l+1th layer And take the utility average as indicator data;

[0074] Step S69: If the l+1th layer is the target layer, then the utility average of the unique indicator of the l+1th layer That is, the evaluation status of the new power system in the kth region; otherwise, repeat steps S61 to S68.

[0075] Compared with the prior art, the present invention has the following beneficial effects: The present invention provides a new comprehensive evaluation method for power systems based on hierarchical evidence reasoning. When comprehensively evaluating the construction of new power systems in different regions, firstly, starting from the subjective expert perspective and objective data analysis, the three methods of hierarchical analysis, CRITIC weight method and Bayesian network reasoning are used to analyze the statistical indicator system of the new power system to obtain the combined weights of each indicator; secondly, based on the indicator weights and the indicator data of each region, the decomposed certainty and uncertainty evidence are integrated in combination with evidence reasoning, which greatly improves the scientific nature of the comprehensive evaluation of the new power system under uncertain information. The application of the proposed new comprehensive evaluation method for power systems based on hierarchical evidence reasoning can effectively integrate subjective and objective weights and avoid the drawbacks of the subjective and objective methods; it can solve the drawbacks of the existing comprehensive evaluation methods that ignore uncertain information and effectively decompose certain and uncertain evidence; it can effectively integrate certain and uncertain evidence, so as to accurately conduct a comprehensive evaluation of the new power system containing uncertain information. BRIEF DESCRIPTION OF THE DRAWINGS

[0076] Figure 1 A schematic diagram of a Bayesian network generated for an embodiment of the present invention.

[0077] Figure 2 This is a flow chart of a novel power system comprehensive evaluation method based on hierarchical evidential reasoning according to an embodiment of the present invention. DETAILED DESCRIPTION

[0078] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0079] It should be noted that the following detailed descriptions are illustrative and are intended to provide further explanation of the present application. Unless otherwise specified, all technical and scientific terms used herein have the same meanings as those commonly understood by those skilled in the art to which the present application belongs.

[0080] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit the exemplary embodiments according to the present application; as used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that when the terms "comprise" and / or "include" are used in this specification, they indicate the presence of features, steps, operations, devices, components and / or their combinations.

[0081] like Figure 1-2 As shown, the present invention is a novel comprehensive evaluation method of power system based on hierarchical evidence reasoning, comprising the following steps:

[0082] Step S1: Select the first level indicator I of the new power system statistical indicator system l ;

[0083] Step S2: Use Satty comparison scale and expert survey method to compare the first level indicator I l Construct a pairwise comparison matrix C; at the same time, construct an indicator data matrix X based on the indicator data of the first layer;

[0084] Step S3: Use the hierarchical analysis method to calculate and analyze the pairwise comparison matrix and output the subjective weight W 1 At the same time, the CRITIC weight method is used to calculate the indicator data matrix X and output the objective weight W 2 ;

[0085] Step S4: Use Bayesian network reasoning to derive the subjective weight W 1 and objective weight W 2 Perform combined weighting and output comprehensive weight W (n) ;

[0086] Step S5: selecting the index data of the first layer of the statistical index system of the new power system in the kth region;

[0087] Step S6: Use the evidential reasoning method to calculate the utility average u of the l+1th layer indicator avg (I l+1 k ), as the indicator data of the l+1th layer indicator;

[0088] Step S7: Determine whether the current region is the last region; if not, then k+1, execute steps S5 to S6; if yes, then execute the next step.

[0089] Step S8: Determine whether the current indicator level is the last level indicator; if not, then l+1, execute steps S1 to S7; if so, output the utility average value u corresponding to all regions avg (A) As a comprehensive evaluation score of new power systems in all regions.

[0090] Specific:

[0091] 1. Select the first level indicator of the new power system statistical indicator system l .

[0092] 2. Using Satty comparison scale and expert survey method to compare the first level indicators I l Construct a pairwise comparison matrix C; at the same time, construct an indicator data matrix X based on the indicator data of the first layer, including the following steps:

[0093] Step 21: Establish a hierarchical structure based on the subordinate relationship between two adjacent layers of indicators: take the statistical indicator data of the new power system in each region as the scheme layer; take the indicator system as the criterion layer, assuming that the indicator system has a total of l max Layer index, then the criterion layer is l max Layer, where all n indicators in the lth layer of the indicator system are y l ={y l (1) ,y l (2) ,...,y l (n)}; The construction status of the new power system is taken as the target layer.

[0094] Step 22: Use the Satty comparison scale to invite researchers to compare the n indicators y in the first layer based on their own understanding. l (n) Score the importance of the indicators (1 to 9 points);

[0095] Step 23: Compare all n indicators in the lth layer pairwise and assign indicator y l ={y l (1) ,y l (2) ,...,y l (n) For the Jth criterion CR in the l+1th layer k The importance of taking two elements at a time and Use c ij express and CR J The comparison results can be used to construct a pairwise comparison judgment matrix (or contrast matrix), using Represents; where Z is the judgment matrix C J The order of is determined by the number of indicators in the lth layer belonging to the kth indicator in the l+1th layer;

[0096] Step 24: Get all n indicators y of the lth layer in all k regions l The indicator data of the first layer are obtained by querying the relevant statistical data of the relevant departments, and the indicator data of the remaining layers are obtained from the indicators of the l-1 layer through the evidence reasoning method; construct the indicator data matrix X = (x ij ) n×k , where x ij (i∈(1,n),j∈(1,k)) represents the data value of indicator i in the new power system statistical indicator system I of the jth region;

[0097] Step 25: Normalize the indicator data matrix X to eliminate the dimension:

[0098] When the indicator is a positive indicator, the normalization method is as follows:

[0099]

[0100] When the indicator is a negative indicator, the normalization method is as follows:

[0101]

[0102] In the formula, represents the minimum value in index i, represents the maximum value in index i;

[0103] Finally, the normalized indicator data matrix X′=(x i ' j ) n×k .

[0104] 3. Use the hierarchical analysis method to calculate and analyze the pairwise comparison matrix and output the subjective weight W 1 At the same time, the CRITIC weight method is used to calculate the indicator data matrix X and output the objective weight W 2 , including the following steps:

[0105] Step 31: Calculate the judgment matrix C J Find the largest eigenvalue λ among the eigenvalues max , calculate the consistency index C·I of the judgment matrix according to formula (3):

[0106]

[0107] Step 32: According to the judgment matrix order Z, query the average random consistency index R·I in Table 2:

[0108] Table 2 Satty consistency index R·I

[0109] Z 1 2 3 4 5 6 7 8 9 10 11 RI 0 0 0.58 0.90 1.12 1.24 1.32 1.41 1.45 1.49 1.51

[0110] Step 33: Calculate the consistency ratio C·R according to formula (4):

[0111]

[0112] When C·R<0.1, it is considered that the judgment matrix C J The consistency is acceptable; otherwise, researchers need to re-conduct pairwise comparisons and make judgments on the matrix C. J Adjust it so that it can pass the consistency test;

[0113] Step 34: According to the judgment matrix C J and the maximum eigenvalue λ max From formula C J ω J,1 =λ max ω J,1 Calculating λ max The corresponding eigenvector ω J,1 , that is, the eigenvector is It is composed of the subjective weights of the Z indicators in the lth layer that belong to the Jth indicator in the l+1th layer.

[0114] Step 35: According to the normalized indicator data matrix X′=(x i ' j ) n×k The variability of the indicator data is calculated by formula (5):

[0115]

[0116] In the formula, represents the average value of index i, x′ ij represents the data value of indicator i in the jth region after normalization, S i represents the standard deviation of index i;

[0117] Step 36: Calculate the conflict of indicator data using formula (6):

[0118]

[0119] In the formula, R i represents the correlation of indicator i, r zirepresents the correlation coefficient between indicator i and indicator z;

[0120] Step 37: Calculate the information content of the indicator:

[0121]

[0122] Among them, C i The larger it is, the more information the indicator i contains, and more weight coefficients should be assigned to it.

[0123] Step 38: Calculate indicator weights:

[0124]

[0125] Repeating steps 35 to 38 in sequence, the objective weights of the Z l-th layer indicators that belong to the J-th indicator in the l+1th layer can be obtained, and together they constitute the objective weight vector

[0126] Step 39: Repeat steps 31 to 38 to obtain the subjective weights and objective weights of all indicators in the first layer of the new power system statistical indicator system.

[0127] 4. Using Bayesian network reasoning to calculate the subjective weights W of all n indicators in the first layer of the new power system statistical indicator system 1 and objective weight W 2 Perform combined weighting and output comprehensive weight W, including the following steps:

[0128] Step 41: Assume that the subjective weight W of the i-th indicator in the l-th layer of the new power system statistical indicator system is 1 (i) The corresponding subjective weight coefficient is l 1 (i) , objective weight W 2 (i) The corresponding objective weight coefficient is l 2 (i) ;

[0129] Step 42: Construct a Bayesian network structure, set the subjective and objective weights of indicator i as parent nodes, and set the comprehensive weight as child nodes, such as Figure 1 shown.

[0130] Step 43: Set the prior probability distribution of each node in the Bayesian network, where the prior probability distribution of parent node 1 is 1 (i) , the prior probability distribution of parent node 2 is 2 (i) ,Using uniform distribution as the prior probability distribution of comprehensive weights, thus determining the complete Bayesian network;

[0131] Step 44: Use the BNT toolbox, then determine the reasoning engine, update the probability distribution of the parent node according to the result to cause, that is, the value of the known child node, and then use the backward reasoning method to perform reverse reasoning;

[0132] Step 45: Convert the Bayesian network into a joint tree: Convert the Bayesian network into a secondary structure, namely, a joint tree, and then perform probability calculation through the message passing process defined on the joint tree;

[0133] Step 46: adopting a tree-shaped joint probability inference algorithm to update the edge probability distribution of the node by passing messages between the joint tree nodes, wherein the evidence of the comprehensive weight is set to 1;

[0134] Step 47: Local update, and finally get the global reasoning result, that is, infer the posterior probability λ of the parent node 1 (i*) and λ 2 (i*) .

[0135] Step 48: Calculate the comprehensive weight W of indicator i (i) :

[0136] W (i) =λ 1 (i*) W 1 (i) +λ 2 (i*) W 2 (i) (9)

[0137] Step 49: Repeat steps 41 to 48 to obtain the comprehensive weights of all indicators in the first layer of the new power system statistical indicator system.

[0138] 5. Select the indicator data of the first level of the statistical indicator system for the new power system in the kth region.

[0139] 6. Combining the indicator data and comprehensive weights of all n indicators in the first layer of the new power system statistical indicator system, the hierarchical evidential reasoning method is used to obtain all indicator data in the first layer + 1, including the following steps:

[0140] Step 61: Construct an evaluation set L with a total of h evaluation levels = {L 1 ,L 2 ,...,L h}, then the evaluation set of all n indicators in the kth region and the lth layer is y k,l ={y k,l (1) ,y k,l (2) ,...,yk,l (n)};

[0141] Step 62: Set the best and worst indicator evaluation matrices respectively and According to formula (10), the indicator evaluation matrix of the indicator i at the lth level of the kth region is obtained based on the indicator data.

[0142]

[0143] Step 63: For the index i of the kth region and the lth level, its index evaluation matrix Combined with the corresponding index weight W (i) Multiply them together to get the basic credibility matrix of the indicator Right now

[0144] Step 64: Combine the weights W (i) And the index evaluation vector According to equations (11) to (13), the uncertainty basic credibility matrix of indicator i is obtained:

[0145]

[0146] Step 65: According to equations (14) to (18), the Z indicator evaluation matrices of the kth region belonging to indicator J in the l+1th layer are merged to calculate the total basic credibility matrix of indicator J: And the total uncertainty basic credibility matrix

[0147]

[0148] In the formula, I(i) represents the comprehensive index after integrating the first i indicators. is the total basic credibility of the unassigned indicator I(i), that is, the total uncertainty basic credibility assigned to indicator I(i), I(i+1) represents the comprehensive indicator after integrating the first i+1 indicators, is the total basic credibility of indicator I(i+1) at evaluation level t;

[0149] Step 66: Fuse the evaluation matrix of the Z indicators in the lth layer to obtain the total basic credibility matrix of the indicator J in the l+1th layer and the total uncertainty basic credibility According to formula (19) and formula (20), the total confidence of the l+1th layer indicator J at each evaluation level is calculated as and the total uncertainty confidence

[0150]

[0151] Step 67: Determine the evaluation level set L = {L 1 ,L 2 ,...,L h The corresponding utility value set is u(L)={u(L 1 ),u(L 2 ),...,u(L h )}, according to formula (21) to formula (23), the utility average value of the l+1th level indicator J in the kth region can be calculated

[0152]

[0153]

[0154] Step 68: Repeat steps 61 to 67 to calculate the utility average of all indicators in the l+1th layer And take the utility average as indicator data;

[0155] Step 69: If the l+1th layer is the target layer, then the utility average of the unique indicator in the l+1th layer That is the evaluation situation of the new power system in the kth region; otherwise, repeat steps 61 to 68.

[0156] 7. Determine whether the current region is the last region; if not, then k+1, execute steps 5 to 6; if yes, then execute the next step.

[0157] 8. Determine whether the current indicator level is the last level indicator; if not, then l+1, execute steps 1 to 7; if so, output the utility average u corresponding to all regions avg (A) As a comprehensive evaluation score of new power systems in all regions.

[0158] The embodiments of the present invention are described in detail above with reference to the accompanying drawings, but the present invention is not limited to the described embodiments. For those skilled in the art, various changes, modifications, substitutions and variations of these embodiments are made without departing from the principles and spirit of the present invention, and still fall within the scope of protection of the present invention.

[0159] The above system and method provided in this embodiment can be stored in a computer-readable storage medium in a coded form and implemented in the form of a computer program, and the basic parameter information required for calculation is input through computer hardware, and the calculation result is output.

[0160] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0161] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0162] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0163] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process in the computer or other programmable device. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0164] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the relevant field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents. Any modification or equivalent replacement that does not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

[0165] This patent is not limited to the above-mentioned optimal implementation mode. Anyone can derive other various forms of new power system statistical indicator system evaluation system based on multi-expert system evidence reasoning under the inspiration of this patent. All equal changes and modifications made according to the scope of the patent application of the present invention should be covered by this patent.

Claims

1. A new comprehensive evaluation method for power systems based on hierarchical evidential reasoning, characterized in that: The steps include: Step S1: Select the first level indicator I of the new power system statistical indicator system l ; Step S2: Use Satty comparison scale and expert survey method to compare the first level indicator I l Construct a pairwise comparison matrix C; at the same time, construct an indicator data matrix X based on the indicator data of the first layer; Step S3: Use the hierarchical analysis method to calculate and analyze the pairwise comparison matrix, output the subjective weight W1, and use the CRITIC weight method to calculate the indicator data matrix X, and output the objective weight W2; Step S4: Use Bayesian network reasoning to combine and weight the subjective weight W1 and the objective weight W2, and output the comprehensive weight W (n) ; Step S5: selecting the index data of the first layer of the statistical index system of the new power system in the kth region; Step S6: Use the evidential reasoning method to calculate the utility average u of the l+1th layer indicator avg (I l+1 k ), as the indicator data of the l+1th layer indicator; Step S7: Determine whether the current region is the last region; if not, then k+1, execute steps S5 to S6; if yes, then execute the next step; Step S8: Determine whether the current indicator level is the last level indicator; if not, then l+1, execute steps S1 to S7; if so, output the utility average value u corresponding to all regions avg (A) As a comprehensive evaluation score of new power systems in all regions.

2. The novel comprehensive evaluation method of power system based on hierarchical evidence reasoning according to claim 1 is characterized in that: The step S2 comprises the following steps: Step S21: Establish a hierarchical structure based on the subordinate relationship of two adjacent layers of indicators: take the statistical indicator data of the new power system in each region as the scheme layer; take the indicator system as the criterion layer, assuming that the indicator system has l max Layer index, then the criterion layer is l max Layer, where all n indicators in the lth layer of the indicator system are y l ={y l (1) ,y l (2) ,...,y l (n) }; The construction of new power systems is taken as the target layer; Step S22: Using the Satty comparison scale, invite researchers to compare the n indicators y in the first layer according to their own understanding. l (n) Score the importance of indicators; Step S23: Compare all n indicators in the first layer in pairs and assign indicator y l ={y l (1) ,y l (2) ,...,y l (n) For the Jth criterion CR in the l+1th layer k The importance of taking two elements at a time and Use c ij express and CR J The comparison results are used to construct a judgment matrix for pairwise comparison or pairwise comparison. Indicates; where Z is the judgment matrix C J The order of is determined by the number of indicators in the lth layer belonging to the kth indicator in the l+1th layer; Step S24: Obtain all n indicators y of the lth layer in all k regions l The indicator data of the first layer is obtained by querying statistical data, and the indicator data of the remaining layers are obtained from the indicators of the l-1 layer through the evidence reasoning method; construct the indicator data matrix X = (x ij ) n×k , where x ij (i∈(1,n),j∈(1,k)) represents the data value of indicator i in the new power system statistical indicator system I of the jth region; n represents the number of indicators; k represents the number of regions; Step S25: normalize the indicator data matrix X to eliminate the dimension: When the indicator is a positive indicator, the normalization method is: When the indicator is a negative indicator, the normalization method is: In the formula, x ij represents the data value of the i-th indicator in the j-th region; x ij ' represents the normalized data value of the ith indicator in the jth region; represents the minimum value in index i, represents the maximum value in index i; Get the normalized indicator data matrix X′=(x i ' j ) n×k .

3. The novel comprehensive evaluation method of power system based on hierarchical evidence reasoning according to claim 1 is characterized in that: The step 3 comprises the following steps: Step S31: Calculate the judgment matrix C J Find the largest eigenvalue λ among the eigenvalues max , calculate the consistency index C·I of the judgment matrix according to formula (3): Step S32: Query the average random consistency index R·I: Step S33: Calculate the consistency ratio C·R according to formula (4): When C·R<0.1, it is considered that the judgment matrix C J The consistency is acceptable; otherwise, researchers need to re-conduct pairwise comparisons and make judgments on the matrix C. J Adjust it so that it can pass the consistency test; Step S34: According to the judgment matrix C J and the maximum eigenvalue λ max From formula C J ω J,1 =λ max ω J,1 Calculating λ max The corresponding eigenvector ω J,1 , that is, the eigenvector is It is composed of the subjective weights of the Z indicators in the lth layer that belong to the Jth indicator in the l+1th layer; Step S35: According to the normalized index data matrix X′=(x i ' j ) n×k The variability of the indicator data is calculated by formula (5): In the formula, represents the average value of index i, x i ' j represents the data value of indicator i in the jth region after normalization, S i represents the variability of indicator i; Step S36: Calculate the conflict of indicator data using formula (6): In the formula, R i represents the correlation of indicator i, r zi represents the correlation coefficient between indicator i and indicator z; Step S37: Calculate the information volume of the index: In the formula, C i Indicates the amount of information contained in indicator i. The more information an indicator has, the more weight coefficients it is assigned. i represents the variability of indicator i; Step S38: Calculate the index weight: Repeat steps S35 to S38 in sequence to obtain the objective weights of the Z l-th layer indicators that belong to the J-th indicator in the l+1-th layer, and together form the objective weight vector Step S39: Repeat steps S31 to S38 to obtain the subjective weights and objective weights of all indicators in the first layer of the new power system statistical indicator system.

4. The novel power system comprehensive evaluation method based on hierarchical evidence reasoning according to claim 1 is characterized in that: The step S4 comprises the following steps: Step S41: Assume that the subjective weight W1 of the i-th indicator in the l-th layer of the new power system statistical indicator system is (i) The corresponding subjective weight coefficient is l1 (i) , objective weight W2 (i) The corresponding objective weight coefficient is l2 (i) ; Step S42: constructing a Bayesian network structure, setting the subjective and objective weights of indicator i as the parent node, and setting the comprehensive weight as the child node; Step S43: Set the prior probability distribution of each node in the Bayesian network, where the prior probability distribution of the first parent node is 1 (i) , the prior probability distribution of the second parent node is 2 (i) ,Using uniform distribution as the prior probability distribution of comprehensive weights, thus determining the complete Bayesian network; Step S44: using the BNT toolbox, then determining the reasoning engine, updating the probability distribution of the parent node according to the process from effect to cause, that is, the value of the known child node, so as to perform reverse reasoning using the backward reasoning method; Step S45: Convert the Bayesian network into a joint tree: Convert the Bayesian network into a secondary structure, i.e., a joint tree, and then perform probability calculation through a message passing process defined on the joint tree; Step S46: using a tree-shaped joint probability inference algorithm to update the edge probability distribution of the node by passing messages between the joint tree nodes, wherein the evidence of the comprehensive weight is set to 1; Step S47: Local update, and finally obtain the global reasoning result, that is, the posterior probability λ1 of the parent node is inferred (i*) and λ2 (i*) ; Step S48: Calculate the comprehensive weight W of index i (i) : W (i) =λ1 (i*) W1 (i) +λ2 (i*) W2 (i) (9) Step S49: Repeat steps S41 to S48 to obtain the comprehensive weights of all indicators in the first layer of the new power system statistical indicator system.

5. The novel comprehensive evaluation method of power system based on hierarchical evidence reasoning according to claim 1 is characterized in that: The step S6 comprises the following steps: Step S61: Construct an evaluation set L with a total of h evaluation levels = {L1, L2, ..., L h }, then the evaluation set of all n indicators in the kth region and the lth layer is y k,l ={y k,l (1) ,y k,l (2) ,...,y k,l (n) }; Step S62: Set the best and worst indicator evaluation matrices respectively and According to formula (10), the indicator evaluation matrix of the indicator i at the lth level of the kth region is obtained based on the indicator data. Step S63: For the index i of the kth region and the lth level, its index evaluation matrix Combined with the corresponding index weight W (i) Multiply them together to get the basic credibility matrix of the indicator Right now Step S64: Combine the weights W (i) And the index evaluation vector According to equations (11) to (13), the uncertainty basic credibility matrix of indicator i is obtained: In the formula, represents the residual uncertainty of the ith indicator in the kth region; It represents the uncertainty after weighted allocation of the ith indicator in the kth region; It represents the confidence of the i-th indicator in the k-th region at the j-th evaluation level; Step S65: According to equations (14) to (18), the Z index evaluation matrices of the kth region belonging to the index J in the l+1th layer are merged to calculate the total basic credibility matrix of the index J: And the total uncertainty basic credibility matrix In the formula, I(i) represents the comprehensive index after integrating the first i indicators. is the total basic credibility of the unassigned indicator I(i), that is, the total uncertainty basic credibility assigned to indicator I(i), I(i+1) represents the comprehensive indicator after integrating the first i+1 indicators, is the total basic credibility of indicator I(i+1) at evaluation level t; Step S66: Fuse the Z index evaluation matrix of the first layer to obtain the total basic credibility matrix of the index J of the first layer + 1 and the total uncertainty basic credibility According to formula (19) and formula (20), the total confidence of the l+1th layer indicator J at each evaluation level is calculated as and the total uncertainty confidence Step S67: Determine the evaluation level set L = {L1, L2, ..., L h The corresponding utility value set is u(L)={u(L1),u(L2),...,u(L h )}, according to formula (21) to formula (23), the utility average value of the l+1th level indicator J in the kth region can be calculated Step S68: Repeat steps S51 to S57 to calculate the utility average of all indicators in the l+1th layer And take the utility average as indicator data; Step S69: If the l+1th layer is the target layer, then the utility average of the unique indicator of the l+1th layer That is, the evaluation status of the new power system in the kth region; otherwise, repeat steps S61 to S68.

Images