Electric vehicle response potential grading method and system and electric vehicle scheduling method participating in power grid frequency modulation and peak regulation

Abstract

The invention discloses an electric vehicle response potential grading method, an electric vehicle response potential grading system and an electric vehicle scheduling method participating in peak regulation of a power grid, and the electric vehicle response potential grading method comprises the following steps: firstly, constructing an electric vehicle demand response potential evaluation system; obtaining a first response evaluation index value and a second response evaluation index value of each of the plurality of electric vehicles based on the constructed system, and then respectively carrying out clustering related to the physical attribute dimension and the economic attribute dimension; correspondingly obtaining a first clustering result of the electric vehicle in the physical attribute dimension and a second clustering result of the electric vehicle in the economic attribute dimension; and finally, carrying out correlation coupling on the clustering result to obtain an electric vehicle comprehensive clustering result, and taking the electric vehicle comprehensive clustering result as an electric vehicle response potential grading scheme. According to the method, accurate hierarchical regulation and control of electric vehicle participation demand response are realized by integrating physical and economic levels.

Description

Technical Field

[0001] The present invention belongs to the technical field of electric vehicles, and in particular relates to a method and system for grading electric vehicle response potential and a method for dispatching electric vehicles participating in power grid frequency and peak regulation. Background Art

[0002] Electric vehicles have attracted much attention in energy conservation, emission reduction and demand response (DR) due to their good environmental protection characteristics and load controllability. As a flexible demand response resource, real-time control of electric vehicles is an important means to solve the peak load regulation of the power grid.

[0003] In the practice of electric vehicles participating in peak load regulation, the assessment of electric vehicle potential is the key. Influenced by the goals of comprehensive promotion and user perception, the grid side or the aggregator side often focuses on the regulation potential represented by the physical level of electric vehicles, such as vehicle battery capacity, access to the grid and peak time, and adopts the same compensation standard in the dispatch of electric vehicles, ignoring the differences in economic incentives between individual electric vehicles, giving high electricity price subsidies to electric vehicles with low regulation potential, resulting in waste, and giving low electricity price subsidies to electric vehicles with high regulation potential, resulting in low enthusiasm. For this reason, some studies set indicators such as user satisfaction to characterize the enthusiasm and willingness of electric vehicles to participate in interaction, but these indicators are often based on assumptions, and indicators such as user satisfaction are somewhat subjective, resulting in this type of method mainly remaining at the research level and difficult to promote in practice.

[0004] As the participation of electric vehicles in grid peak load regulation becomes normalized, when regulating a large number of electric vehicle users, how to comprehensively evaluate the charging and discharging response potential of electric vehicles by combining the objective response capability at the physical level and the subjective response willingness at the economic level, and based on this, construct a regulation strategy for cluster electric vehicles to participate in grid peak load regulation, is of great significance for achieving the normalization of electric vehicles participating in grid peak load regulation. Summary of the invention

[0005] In order to solve the problems existing in the prior art, the present invention proposes a method and system for grading the response potential of electric vehicles and a method for dispatching electric vehicles participating in the frequency and peak regulation of a power grid.

[0006] The technical solution of the present invention is as follows:

[0007] A method for grading electric vehicle response potential, the method comprising:

[0008] Constructing an electric vehicle demand response potential evaluation system, wherein the evaluation indicators of the electric vehicle demand response potential evaluation system include a first response evaluation indicator for reflecting the physical attributes of the electric vehicle and a second response evaluation indicator for reflecting the economic attributes of the electric vehicle;

[0009] Based on the historical operation data of the electric vehicle, under the electric vehicle demand response potential evaluation system, obtaining a first response evaluation index value and a second response evaluation index value of each of the plurality of electric vehicles;

[0010] Based on the first response evaluation index value and the second response evaluation index value, clustering is performed in relation to the physical attribute dimension and the economic attribute dimension, respectively, and a first clustering result in relation to the physical attribute dimension and a second clustering result in relation to the economic attribute dimension of the electric vehicle are correspondingly obtained;

[0011] The first clustering result and the second clustering result are correlated and coupled to obtain a comprehensive clustering result of electric vehicles as a grading scheme for the response potential of electric vehicles.

[0012] Furthermore, the first response evaluation index includes vehicle battery capacity, initial charging state of charge, expected state of charge and grid access distance from peak time; the second response evaluation index includes peak-to-valley charging cost ratio and compensation price sensitivity.

[0013] Furthermore, the vehicle battery capacity is the battery capacity of the user's vehicle, and its expression is:

[0014] x1=Q m ;

[0015] In the formula, Q m Indicates the vehicle battery capacity;

[0016] The initial charging state of charge is the battery power percentage of the electric vehicle when it is connected to the grid, and its expression is:

[0017]

[0018] Where x2 is the initial state of charge; Q s and Q m They are the remaining capacity and rated capacity of the electric vehicle respectively;

[0019] The expected state of charge is the battery power percentage that the electric vehicle is expected to reach when charging, and its expression is:

[0020]

[0021] Where x3 is the initial state of charge; Q y and Q m They are the expected capacity and rated capacity of the electric vehicle at the end of charging;

[0022] The time between grid access and peak load is the time interval between the time when the electric vehicle grid accesses and the start of the peak load regulation period, and its expression is:

[0023]

[0024] Where x4 is the distance between the peak and the grid time of electric vehicles; t r is the time when the electric vehicle enters the grid; t f The peak time starts.

[0025] The peak-to-valley charging cost ratio is the ratio of the charging cost of the electric vehicle during the peak period to the charging cost during the valley period, and its expression is:

[0026]

[0027] Where x5 is the peak-to-valley charging cost ratio of electric vehicles; C f and C g They are peak-time charging fee and valley-time charging fee respectively;

[0028] The compensation price sensitivity is the impact of compensation price changes on the demand for electric vehicles, and its expression is:

[0029]

[0030] Where x6 is the compensation price sensitivity; X and C b are the percentage change in demand for electric vehicle charging and the percentage change in compensation price, respectively.

[0031] Furthermore, the specific method for obtaining the first response evaluation index value and the second response evaluation index value of each electric vehicle in the electric vehicle demand response potential evaluation system based on the electric vehicle historical operation data includes:

[0032] Optimization of indicator weights using the CRITIC method:

[0033] For m electric vehicle samples and n response evaluation indicators, first obtain the initial classification indicator matrix D. The value x in row i and column j of the initial classification indicator matrix D is ij represents the jth response evaluation index value of the i-th electric vehicle sample, and the element after the standardized processing of the matrix is ​​x ij ,

[0034] For positive indicators:

[0035]

[0036] For negative indicators

[0037]

[0038] The variability of the response evaluation index value is calculated, where the calculation formula for the variability of the j-th response evaluation index value is as follows:

[0039]

[0040] In the formula, x ij is the element in the matrix after normalization, x j is the average value of the j-th response evaluation indicator.

[0041] Then calculate the conflict between the response evaluation indicators: Let the conflict between the jth response evaluation indicator value and the remaining indicators be f j ,but

[0042]

[0043] Among them, r ij represents the correlation coefficient between the i-th response evaluation index value and the j-th response evaluation index value;

[0044] The response evaluation index information volume is calculated, where the calculation formula of the response evaluation index information volume of the j-th response evaluation index is as follows:

[0045] C j =σ j f j

[0046] Calculate the weight of the response evaluation index, where the calculation formula for the weight of the j-th response evaluation index is as follows:

[0047]

[0048] Calculate the user response potential index. The calculation formula for the i-th response evaluation index value is as follows:

[0049]

[0050] Furthermore, the specific method of performing clustering related to the physical attribute dimension and the economic attribute dimension respectively based on the first response evaluation index value and the second response evaluation index value, and correspondingly obtaining the first clustering result of the electric vehicle in the physical attribute dimension and the second clustering result in the economic attribute dimension includes:

[0051] 1) Assuming that the data volume of all first response evaluation indicators reflecting the physical properties of electric vehicles and the second response evaluation indicators reflecting the economic properties of electric vehicles is n, the number of randomly selected initial clustering centers of the data is K;

[0052] 2) Calculate the Euclidean distance between each sample data and the initial cluster center;

[0053] 3) Calculate the error square sum function J c ;

[0054] 4) Calculate J cWhether the distance between two iterations is less than the set threshold, if so, the algorithm ends, otherwise, it continues to iterate, calculate the cluster center, and obtain new K cluster centers;

[0055] 5) Return to step 2) until the error square sum function J c is the minimum, or meets the termination condition.

[0056] Furthermore, the method of performing correlation coupling between the first clustering result and the second clustering result to obtain a comprehensive clustering result of electric vehicles as a specific method of the electric vehicle response potential classification scheme includes:

[0057] Let the first clustering result be expressed as in, are the electric vehicle sets belonging to the 1st cluster center, ..., n1th cluster center in the first clustering result respectively;

[0058] And let the second clustering result be expressed as in, are respectively the electric vehicle sets belonging to the 1st cluster center, ..., n2th cluster center in the second clustering result;

[0059] Based on the first clustering result and the second clustering result, a correlation coupling matrix is ​​constructed, and each element in the correlation coupling matrix corresponds to a potential level of electric vehicle response; wherein the correlation coupling matrix is ​​expressed as:

[0060]

[0061] In the formula, Ω 11 ,...,Ω 1n2 The electric vehicle sets that belong to the first cluster center in the first clustering result and the first cluster center in the second clustering result are respectively, ..., the electric vehicle sets that belong to the first cluster center in the first clustering result and the n2th cluster center in the second clustering result;

[0062] Ω 21 ,...,Ω 2n2 The electric vehicle sets that belong to the second cluster center in the first clustering result and the first cluster center in the second clustering result are respectively, ..., the electric vehicle sets that belong to the second cluster center in the first clustering result and the n2th cluster center in the second clustering result;

[0063] Ω n11 ,...,Ω n1n2They are respectively a set of electric vehicles that belong to the n1th cluster center in the first clustering result and the 1st cluster center in the second clustering result, ..., a set of electric vehicles that belong to the n1th cluster center in the first clustering result and the n2th cluster center in the second clustering result.

[0064] An electric vehicle response potential grading system includes a system building module, an evaluation module, a clustering module and a fusion module;

[0065] The system construction module is used to construct an electric vehicle demand response potential evaluation system, wherein the evaluation indicators of the electric vehicle demand response potential evaluation system include a first response evaluation indicator for reflecting the physical properties of the electric vehicle and a second response evaluation indicator for reflecting the economic properties of the electric vehicle;

[0066] The evaluation module is used to obtain a first response evaluation index value and a second response evaluation index value of each electric vehicle among a plurality of electric vehicles under the electric vehicle demand response potential evaluation system based on the electric vehicle historical operation data;

[0067] The clustering module is used to perform clustering related to the physical attribute dimension and the economic attribute dimension respectively based on the first response evaluation index value and the second response evaluation index value, and correspondingly obtain a first clustering result of the electric vehicle in the physical attribute dimension and a second clustering result in the economic attribute dimension;

[0068] The fusion module is used to perform correlation coupling on the first clustering result and the second clustering result to obtain a comprehensive clustering result of electric vehicles as a grading scheme for the response potential of electric vehicles.

[0069] An electronic device comprises a memory and a processor, wherein the memory stores a computer program, and the processor is used to call and run the computer program stored in the memory to execute any of the methods described above.

[0070] A computer-readable storage medium stores a computer program, wherein the computer program implements the steps of any of the above methods when executed by a processor.

[0071] A method for dispatching electric vehicles participating in power grid peak load regulation comprises: adopting any of the above electric vehicle response potential classification methods to classify the response potential of target electric vehicles, and dispatching the target electric vehicles for peak load regulation according to the response potential classification results.

[0072] Compared with the prior art, the present invention has the following beneficial effects:

[0073] The present invention proposes a method and system for grading the response potential of electric vehicles. The method first constructs an electric vehicle demand response potential evaluation system, obtains the first response evaluation index value and the second response evaluation index value of each electric vehicle among a number of electric vehicles based on the constructed system, and then performs clustering related to the physical attribute dimension and the economic attribute dimension respectively, and obtains the first clustering result of the electric vehicle in the physical attribute dimension and the second clustering result in the economic attribute dimension; finally, the clustering results are coupled by correlation to obtain the comprehensive clustering result of the electric vehicle as the electric vehicle response potential grading scheme. The method realizes precise graded regulation of electric vehicles participating in demand response by integrating physical and economic levels.

[0074] The present invention proposes a method for dispatching electric vehicles that participate in power grid frequency and peak regulation. The method first adopts the electric vehicle response potential grading method of the present invention to grade the response potential of target electric vehicles, and dispatches the target electric vehicles according to the response potential grading results. The dispatching method targets and screens matching electric vehicles to participate in power grid operation under the premise of fully considering the multi-dimensional needs of electric vehicles, which is conducive to improving the implementation effect of demand response, providing an effective solution for the implementation of electric vehicles participating in demand response, and thus supporting the stable operation of the power system. BRIEF DESCRIPTION OF THE DRAWINGS

[0075] Figure 1 It is a flow chart of the electric vehicle response potential classification method in the embodiment;

[0076] Figure 2 This is a first clustering result diagram on the physical attribute dimension in the embodiment;

[0077] Figure 3 This is a second clustering result diagram of the economic attribute dimension in the embodiment. DETAILED DESCRIPTION

[0078] The present invention is further explained below in conjunction with the accompanying drawings and specific embodiments. It should be understood that these embodiments are only used to illustrate the present invention and are not used to limit the scope of the present invention. After reading the present invention, various equivalent modifications of the present invention by those skilled in the art all fall within the scope defined by the claims attached to this application.

[0079] Embodiment 1:

[0080] A method for grading electric vehicle response potential of the present invention is as follows: Figure 1 As shown, the method includes:

[0081] Construct an electric vehicle demand response potential assessment system. The evaluation indicators of the electric vehicle demand response potential assessment system include a first response assessment indicator for reflecting the physical attributes of electric vehicles and a second response assessment indicator for reflecting the economic attributes of electric vehicles.

[0082] Based on the historical operation data of electric vehicles (the historical operation data can be retrieved from the electricity consumption information collection system or the electric vehicle management platform), under the electric vehicle demand response potential evaluation system, the first response evaluation index value and the second response evaluation index value of each electric vehicle among a number of electric vehicles are obtained;

[0083] Based on the first response evaluation index value and the second response evaluation index value, clustering is performed in relation to the physical attribute dimension and the economic attribute dimension, respectively, and a first clustering result of the electric vehicle in the physical attribute dimension and a second clustering result in the economic attribute dimension are obtained accordingly;

[0084] The first clustering results and the second clustering results are correlated and coupled to obtain the comprehensive clustering results of electric vehicles, which serve as the grading scheme for the response potential of electric vehicles.

[0085] Embodiment 2:

[0086] This embodiment is further designed on the basis of the first embodiment in that the first response evaluation index includes the vehicle battery capacity, the initial charging state of charge, the expected state of charge and the distance from the peak to the grid access time; the second response evaluation index includes the peak-to-valley charging cost ratio and the compensation price sensitivity.

[0087] Embodiment three:

[0088] This embodiment is further designed on the basis of the second embodiment in that the vehicle battery capacity in this example is the battery capacity of the user's vehicle, reflecting the potential scale level of the user's vehicle dispatchable object, and its expression is:

[0089] x1=Q m ;

[0090] In the formula, Q m Indicates the vehicle battery capacity;

[0091] The initial charging state of charge is the battery power percentage of the electric vehicle when it is connected to the grid, reflecting the charging demand of the electric vehicle. Its expression is:

[0092]

[0093] Where x2 is the initial state of charge; Q s and Q m They are the remaining capacity and rated capacity of the electric vehicle respectively;

[0094] The expected state of charge is the percentage of battery power that the electric vehicle hopes to achieve when charging, reflecting the user's expected charging level. Its expression is:

[0095]

[0096] Where x3 is the initial state of charge; Q y and Q m They are the expected capacity and rated capacity of the electric vehicle at the end of charging;

[0097] The time between grid access and peak load is the time interval between the time when electric vehicles grid access and the start of peak load regulation, which reflects the regulation time margin. The closer to the start of peak load regulation, the smaller the regulation willingness and regulation capacity of electric vehicles; vice versa. Its expression is:

[0098]

[0099] Where x4 is the distance between the peak and the grid time of electric vehicles; t r is the time when the electric vehicle enters the grid; t f The peak time starts.

[0100] The peak-to-valley charging cost ratio is the ratio of the charging cost of electric vehicles during peak hours to the charging cost during valley hours, and its expression is:

[0101]

[0102] Where x5 is the peak-to-valley charging cost ratio of electric vehicles; C f and C g They are peak-time charging fee and valley-time charging fee respectively;

[0103] The compensation price sensitivity is the impact of compensation price changes on the demand for electric vehicles, and its expression is:

[0104]

[0105] Where x6 is the compensation price sensitivity; X and C b are the percentage change in demand for electric vehicle charging and the percentage change in compensation price, respectively.

[0106] Embodiment 4:

[0107] This embodiment is further designed on the basis of the first embodiment in that, in this embodiment, based on the historical operation data of electric vehicles, under the electric vehicle demand response potential evaluation system, the specific method for obtaining the first response evaluation index value and the second response evaluation index value of each electric vehicle among a plurality of electric vehicles includes:

[0108] Optimization of indicator weights using the CRITIC method:

[0109] For m electric vehicle samples (or “EV samples”) and n response evaluation indicators (first / second response evaluation indicators), first obtain the initial classification indicator matrix D. The value x in row i and column j of the initial classification indicator matrix D is ijrepresents the jth response evaluation index value of the i-th electric vehicle sample, and the element after the standardized processing of the matrix is ​​x ij ,

[0110] For positive indicators:

[0111]

[0112] For negative indicators

[0113]

[0114] In the present invention, the vehicle battery capacity and the initial charging state of charge are positive indicators, the expected state of charge and the distance from the peak to the grid are negative indicators, and the peak-to-valley charging cost ratio and the compensation price sensitivity are positive indicators.

[0115] The variability of the response evaluation index value is calculated, where the calculation formula for the variability of the j-th response evaluation index value (first / second response evaluation index) is as follows:

[0116]

[0117] In the formula, x ij is the element in the matrix after normalization, x j is the average value of the j-th response evaluation indicator.

[0118] Then calculate the conflict between the response evaluation indicators: the conflict reflects the correlation between different indicators. If there is a significant positive correlation, the smaller the conflict value is. Let the contradiction between the jth response evaluation indicator value and the remaining indicators (i.e., conflict) be f j ,but

[0119]

[0120] Among them, r ij represents the correlation coefficient between the i-th response evaluation index value and the j-th response evaluation index value; the correlation coefficient represents the conflict between the response evaluation indexes, and the Pearson correlation coefficient is used in the present invention;

[0121] The Pearson correlation coefficient is defined as the quotient of the covariance and standard deviation between two variables:

[0122]

[0123] The above formula defines the overall correlation coefficient, which is usually represented by the Greek lowercase letter ρ. By estimating the covariance and standard deviation of the sample, we can get the Pearson correlation coefficient, which is usually represented by the English lowercase letter r:

[0124]

[0125] r can also be expressed by (X i ,Y i ) The standard score mean estimate of the sample points is obtained, and the expression equivalent to the above formula is obtained:

[0126]

[0127] in and σ X They are respectively i The standard score of the sample, the sample mean, and the sample standard deviation.

[0128] ρ x,y The value is between -1 and +1. The larger the absolute value, the stronger the correlation (if it is a positive number, it indicates that the two variables are positively correlated; if it is a negative number, it indicates that the two variables are negatively correlated).

[0129] In order to reflect the importance of the index on the response potential, the information volume of the response evaluation index is calculated. The calculation formula of the information volume of the response evaluation index of the jth response evaluation index is as follows:

[0130] C j =σ j f j

[0131] Calculate the weight of the response evaluation index, C j The larger the value, the more information the j-th response evaluation indicator carries and the greater its importance, that is, the j-th indicator weight w j The larger the value, the weight of the j-th response evaluation index is calculated as follows:

[0132]

[0133] It can be seen that the greater the information carrying capacity, the greater the weight.

[0134] Calculate the user response potential index. The calculation formula for the i-th response evaluation index value is as follows:

[0135]

[0136] Embodiment five:

[0137] This embodiment is further designed on the basis of the first embodiment in that, in this embodiment, clustering related to the physical attribute dimension and the economic attribute dimension is performed respectively based on the first response evaluation index value and the second response evaluation index value, and the specific method for obtaining the first clustering result of the electric vehicle in the physical attribute dimension and the second clustering result in the economic attribute dimension includes:

[0138] The K-means algorithm is a distance-based clustering algorithm. Its core idea is to divide the data into K clusters in an iterative manner so that the data points in each cluster are as close to the center of the cluster as possible, while the data points between different clusters are as far away as possible. The principle of the K-means clustering algorithm is as follows:

[0139] Define the objective function of the k-means clustering algorithm:

[0140]

[0141] In the formula, r ik It means that when the sample x i It is 1 when it is classified into cluster K, otherwise it is 0; u k represents the mean vector of cluster k;

[0142] The objective function describes to some extent the closeness of the samples in the cluster around the cluster mean vector. The smaller the J value, the higher the similarity of the samples in the cluster. Minimizing the objective function is an NP-hard problem. K-means clustering uses the EM algorithm to optimize the model. Initialize the mean vector of K clusters, that is, u k is a constant, find r when J is minimized ik When a data point is divided into the cluster closest to the data point, the objective function J is minimized. ik , find the corresponding u when minimizing J k .

[0143] Let the objective function J be u k The partial derivative of is equal to 0:

[0144]

[0145] have to:

[0146]

[0147] u k The cluster center is equal to the mean of the samples in the cluster;

[0148] The specific steps of the k-means algorithm are as follows:

[0149] 1) Assuming that the data volume of all first response evaluation indicators reflecting the physical properties of electric vehicles and the second response evaluation indicators reflecting the economic properties of electric vehicles is n, the number of randomly selected initial clustering centers of the data is K;

[0150] 2) Calculate the Euclidean distance between each sample data and the initial cluster center;

[0151] 3) Calculate the error square sum function J c ;

[0152] 4) Calculate J c Whether the distance between two iterations is less than the set threshold, if so, the algorithm ends, otherwise, it continues to iterate, calculate the cluster center, and obtain new K cluster centers;

[0153] 5) Return to step 2) until the error square sum function J c is the minimum, or meets the termination condition.

[0154] The K-means clustering algorithm is used to perform cluster analysis on automobile user behavior. The results of cluster analysis are closely related to the K value. The setting of the K value is subject to too much human intervention, so it is necessary to comprehensively consider the clustering effect under different K values. The effects of dividing the sample data into 2 categories (high response potential, low response potential), 3 categories (high response potential, medium response potential, low response potential) and 4 categories (high response potential, medium response potential, lower response potential, low response potential) are considered respectively, and finally the clustering results under different K values ​​are obtained.

[0155] Embodiment six:

[0156] This embodiment is further designed on the basis of the first embodiment in that the first clustering result and the second clustering result are coupled by correlation in this embodiment to obtain a comprehensive clustering result of electric vehicles, which is used as a specific method of the electric vehicle response potential classification scheme, including:

[0157] Let the first clustering result be expressed as in, are the electric vehicle sets belonging to the 1st cluster center, ..., n1th cluster center in the first clustering result respectively;

[0158] And let the second clustering result be expressed as in, are respectively the electric vehicle sets belonging to the 1st cluster center, ..., n2th cluster center in the second clustering result;

[0159] Based on the first clustering results and the second clustering results, a correlation coupling matrix is ​​constructed, and each element in the correlation coupling matrix corresponds to an electric vehicle response potential level; wherein the correlation coupling matrix is ​​expressed as:

[0160]

[0161] In the formula, Ω 11 ,...,Ω 1n2The electric vehicle sets that belong to the first cluster center in the first clustering result and the first cluster center in the second clustering result are respectively, ..., the electric vehicle sets that belong to the first cluster center in the first clustering result and the n2th cluster center in the second clustering result;

[0162] Ω 21 ,...,Ω 2n2 The electric vehicle sets that belong to the second cluster center in the first clustering result and the first cluster center in the second clustering result are respectively, ..., the electric vehicle sets that belong to the second cluster center in the first clustering result and the n2th cluster center in the second clustering result;

[0163] Ω n11 ,...,Ω n1n2 They are respectively a set of electric vehicles that belong to the n1th cluster center in the first clustering result and the 1st cluster center in the second clustering result, ..., a set of electric vehicles that belong to the n1th cluster center in the first clustering result and the n2th cluster center in the second clustering result.

[0164] Embodiment seven:

[0165] An electric vehicle response potential grading system of the present invention includes a system construction module, an evaluation module, a clustering module and a fusion module;

[0166] A system construction module, used to construct an electric vehicle demand response potential evaluation system, wherein the evaluation indicators of the electric vehicle demand response potential evaluation system include a first response evaluation indicator for reflecting the physical attributes of electric vehicles and a second response evaluation indicator for reflecting the economic attributes of electric vehicles;

[0167] An evaluation module, for obtaining a first response evaluation index value and a second response evaluation index value of each of a plurality of electric vehicles under an electric vehicle demand response potential evaluation system based on historical operation data of the electric vehicle;

[0168] A clustering module, for performing clustering related to the physical attribute dimension and the economic attribute dimension respectively based on the first response evaluation index value and the second response evaluation index value, and correspondingly obtaining a first clustering result of the electric vehicle in the physical attribute dimension and a second clustering result in the economic attribute dimension;

[0169] The fusion module is used to perform correlation coupling on the first clustering result and the second clustering result to obtain the comprehensive clustering result of electric vehicles as a grading scheme for the response potential of electric vehicles.

[0170] Embodiment eight:

[0171] An electronic device comprises a memory and a processor, wherein the memory stores a computer program, and the processor is used to call and run the computer program stored in the memory to execute any of the methods described above.

[0172] A computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of any of the above methods are implemented.

[0173] Embodiment nine:

[0174] A method for dispatching electric vehicles participating in power grid peak regulation of the present invention comprises:

[0175] The electric vehicle response potential classification method of any one of the first to sixth embodiments is used to classify the response potential of the target electric vehicle, and the target electric vehicle is peak-shifted according to the response potential classification result.

[0176] In this example, the electric vehicle response potential classification method of the present invention is simulated and implemented on 50 electric vehicles. The statistical characteristics of the 50 electric vehicles (hereinafter referred to as "users") are shown in Table 1 below.

[0177] Table 1 Statistical characteristics of users

[0178]

[0179]

[0180] In this example, the CRITIC method is used to optimize the indicator weights. The optimized weights are shown in Tables 2 and 3.

[0181] Table 2 Weights of first response evaluation indicators (physical property indicators) after optimization by CRITIC method

[0182] Battery capacity Initial charging state of charge Expected state of charge Time from network access to peak hours 0.2212 0.2591 0.2316 0.2881

[0183] Table 3 Weights of the second response evaluation index (economic characteristic index) after CRITIC method optimization

[0184] Peak-valley charging cost ratio Compensating for price sensitivity 0.4518 0.5482

[0185] According to the data in Table 1 and the index weights in Table 2, the first response evaluation index values ​​(physical property response potential index) of 50 users are calculated as shown in Table 4.

[0186] Table 4 User physical characteristics response potential index

[0187] user Physical Property Response Potential Index user Physical Property Response Potential Index user Physical Property Response Potential Index 1 41.4522 21 85.6826 41 50.1213 2 69.6395 22 91.8763 42 79.6125 3 73.5995 23 67.6195 43 70.6579 4 79.6374 24 87.4575 44 71.4325 5 88.5004 25 76.1323 45 77.9985 6 59.1334 26 85.4046 46 76.8875 7 50.3348 27 85.7236 47 84.4952 8 85.272 28 93.8209 48 59.8565 9 67.8759 29 58.1159 49 46.5244 10 91.0584 30 86.5173 50 94.9818 11 25.325 31 59.1946 12 61.1462 32 85.9584 13 78.9494 33 86.2822 14 94.8888 34 80.9794 15 86.1903 35 92.176 16 96.1371 36 70.9603 17 77.6252 37 99.965 18 66.4143 38 58.2405 19 62.9468 39 48.7374 20 89.4456 40 50.9862

[0188] According to the data in Table 1 and the indicator weights in Table 3, the second response evaluation index values ​​(economic preference response potential index) of 50 users are calculated as shown in Table 5.

[0189] Table 4 User economic preference response potential index

[0190]

[0191]

[0192] The K-means clustering algorithm is used to perform clustering related to the physical attribute dimension and the economic attribute dimension respectively, and the first clustering result of electric vehicles in the physical attribute dimension and the second clustering result in the economic attribute dimension are obtained. The clustering results are as follows: Figure 2 , Figure 3 shown.

[0193] Considering the correlation between the physical and economic systems of electric vehicles, the first clustering results and the second clustering results are correlated and coupled to obtain the comprehensive clustering results of electric vehicles as a classification scheme for the response potential of electric vehicles, as shown in Table 5.

[0194] Table 5 Electric vehicle response potential level classification table

[0195] High physical dimension In the physical dimension Low physical dimension High economic dimension 10,15,16,20,28,50 3,13,23,25,43 40,41,48,49 In the economic dimension 5,21,22,27,32,35,37 2,17,18,44,45 7,11,29,31,38 Low economic dimension 8,14,24,26,30,33,47 4,9,12,19.34,36,42,46 1,6,39

[0196] The above description is only a specific implementation of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by a technician familiar with the technical field within the technical scope disclosed by the present invention should be covered by the protection scope of the present invention.

Claims

1. A method for grading electric vehicle response potential, characterized in that: The method comprises: Constructing an electric vehicle demand response potential evaluation system, wherein the evaluation indicators of the electric vehicle demand response potential evaluation system include a first response evaluation indicator for reflecting the physical attributes of the electric vehicle and a second response evaluation indicator for reflecting the economic attributes of the electric vehicle; Based on the historical operation data of the electric vehicle, under the electric vehicle demand response potential evaluation system, obtaining a first response evaluation index value and a second response evaluation index value of each of the plurality of electric vehicles; Based on the first response evaluation index value and the second response evaluation index value, clustering is performed in relation to the physical attribute dimension and the economic attribute dimension, respectively, and a first clustering result in relation to the physical attribute dimension and a second clustering result in relation to the economic attribute dimension of the electric vehicle are correspondingly obtained; The first clustering result and the second clustering result are correlated and coupled to obtain a comprehensive clustering result of electric vehicles as a grading scheme for the response potential of electric vehicles.

2. The electric vehicle response potential classification method according to claim 1, characterized in that: The first response evaluation index includes vehicle battery capacity, initial charging state of charge, expected state of charge and grid access distance from peak time; the second response evaluation index includes peak-to-valley charging cost ratio and compensation price sensitivity.

3. The electric vehicle response potential classification method according to claim 2, characterized in that: The vehicle battery capacity is the battery capacity of the user's vehicle, and its expression is: x1=Q m ; In the formula, Q m Indicates the vehicle battery capacity; The initial charging state of charge is the battery power percentage of the electric vehicle when it is connected to the grid, and its expression is: Where x2 is the initial state of charge; Q s and Q m They are the remaining capacity and rated capacity of the electric vehicle respectively; The expected state of charge is the battery power percentage that the electric vehicle is expected to reach when charging, and its expression is: Where x3 is the initial state of charge; Q y and Q m They are the expected capacity and rated capacity of the electric vehicle at the end of charging; The time between grid access and peak load is the time interval between the time when the electric vehicle grid accesses and the start of the peak load regulation period, and its expression is: Where x4 is the distance between the peak and the grid time of electric vehicles; t r is the time when the electric vehicle enters the grid; t f The peak time starts. The peak-to-valley charging cost ratio is the ratio of the charging cost of the electric vehicle during the peak period to the charging cost during the valley period, and its expression is: Where x5 is the peak-to-valley charging cost ratio of electric vehicles; C f and C g They are peak-time charging fee and valley-time charging fee respectively; The compensation price sensitivity is the impact of compensation price changes on the demand for electric vehicles, and its expression is: Where x6 is the compensation price sensitivity; X and C b are the percentage change in demand for electric vehicle charging and the percentage change in compensation price, respectively.

4. The electric vehicle response potential classification method according to claim 1, characterized in that: The specific method for obtaining the first response evaluation index value and the second response evaluation index value of each electric vehicle in the electric vehicle demand response potential evaluation system based on the electric vehicle historical operation data includes: Optimization of indicator weights using the CRITIC method: For m electric vehicle samples and n response evaluation indicators, first obtain the initial classification indicator matrix D. The value x in row i and column j of the initial classification indicator matrix D is ij represents the jth response evaluation index value of the i-th electric vehicle sample, and the element after the standardized processing of the matrix is ​​x ij , For positive indicators: For negative indicators The variability of the response evaluation index value is calculated, where the calculation formula for the variability of the j-th response evaluation index value is as follows: In the formula, x ij is the element in the matrix after normalization, x j is the average value of the j-th response evaluation index; Then calculate the conflict between the response evaluation indicators: Let the conflict between the jth response evaluation indicator value and the remaining indicators be f j ,but Among them, r ij represents the correlation coefficient between the i-th response evaluation index value and the j-th response evaluation index value; The response evaluation index information volume is calculated, where the calculation formula of the response evaluation index information volume of the j-th response evaluation index is as follows: C j =σ j f j Calculate the weight of the response evaluation index, where the calculation formula for the weight of the j-th response evaluation index is as follows: Calculate the user response potential index. The calculation formula for the i-th response evaluation index value is as follows:

5. The electric vehicle response potential classification method according to claim 1, characterized in that: The specific method of performing clustering related to the physical attribute dimension and the economic attribute dimension based on the first response evaluation index value and the second response evaluation index value, and correspondingly obtaining the first clustering result of the electric vehicle in the physical attribute dimension and the second clustering result in the economic attribute dimension includes: 1) Let the data volume of the response evaluation index be n, and the number of randomly selected initial clustering centers of the data be K; 2) Calculate the Euclidean distance between each sample data and the initial cluster center; 3) Calculate the error square sum function J c ; 4) Calculate J c Whether the distance between two iterations is less than the set threshold, if so, the algorithm ends, otherwise, it continues to iterate, calculate the cluster center, and obtain new K cluster centers; 5) Return to step 2) until the error square sum function J c is the minimum, or meets the termination condition.

6. The electric vehicle response potential classification method according to claim 1, characterized in that: The method of performing correlation coupling between the first clustering result and the second clustering result to obtain a comprehensive clustering result of electric vehicles as a specific method of the electric vehicle response potential classification scheme includes: Let the first clustering result be expressed as in, are the electric vehicle sets belonging to the 1st cluster center, ..., n1th cluster center in the first clustering result respectively; And let the second clustering result be expressed as in, are respectively the electric vehicle sets belonging to the 1st cluster center, ..., n2th cluster center in the second clustering result; Based on the first clustering result and the second clustering result, a correlation coupling matrix is ​​constructed, and each element in the correlation coupling matrix corresponds to a potential level of electric vehicle response; wherein the correlation coupling matrix is ​​expressed as: In the formula, Ω 11 ,...,Ω 1n2 The electric vehicle sets that belong to the first cluster center in the first clustering result and the first cluster center in the second clustering result are respectively, ..., the electric vehicle sets that belong to the first cluster center in the first clustering result and the n2th cluster center in the second clustering result; Ω 21 ,...,Ω 2n2 The electric vehicle sets that belong to the second cluster center in the first clustering result and the first cluster center in the second clustering result are respectively, ..., the electric vehicle sets that belong to the second cluster center in the first clustering result and the n2th cluster center in the second clustering result; Ω n11 ,...,Ω n1n2 They are respectively a set of electric vehicles that belong to the n1th cluster center in the first clustering result and the 1st cluster center in the second clustering result, ..., a set of electric vehicles that belong to the n1th cluster center in the first clustering result and the n2th cluster center in the second clustering result.

7. An electric vehicle response potential grading system, characterized in that: It includes system construction module, evaluation module, clustering module and fusion module; The system construction module is used to construct an electric vehicle demand response potential evaluation system, wherein the evaluation indicators of the electric vehicle demand response potential evaluation system include a first response evaluation indicator for reflecting the physical properties of the electric vehicle and a second response evaluation indicator for reflecting the economic properties of the electric vehicle; The evaluation module is used to obtain a first response evaluation index value and a second response evaluation index value of each electric vehicle among a plurality of electric vehicles under the electric vehicle demand response potential evaluation system based on the electric vehicle historical operation data; The clustering module is used to perform clustering related to the physical attribute dimension and the economic attribute dimension respectively based on the first response evaluation index value and the second response evaluation index value, and correspondingly obtain a first clustering result of the electric vehicle in the physical attribute dimension and a second clustering result in the economic attribute dimension; The fusion module is used to perform correlation coupling on the first clustering result and the second clustering result to obtain a comprehensive clustering result of electric vehicles as a grading scheme for the response potential of electric vehicles.

8. An electronic device, characterized in that: The electronic device comprises a memory and a processor, wherein the memory stores a computer program, and the processor is configured to call and run the computer program stored in the memory to execute the method according to any one of claims 1 to 5.

9. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 5 are implemented.

10. A method for dispatching electric vehicles to participate in power grid peak load regulation, characterized in that: include: The electric vehicle response potential classification method according to any one of claims 1 to 5 is used to classify the response potential of target electric vehicles, and the target electric vehicles are peak-shifted according to the response potential classification results.

Images