Method for selecting and sizing of charging station based on entropy weight method and queuing theory

Abstract

This invention discloses a site selection and capacity determination method for charging and battery swapping stations based on entropy weighting and queuing theory. By acquiring and processing multi-source spatial data to construct an object dataset, a candidate index set is formed, and a retained index set is obtained after redundancy detection. After defining upper and lower quantiles, proportional coefficients and annual entropy weights are calculated, combined with preset time smoothing weights to obtain smoothing weights. Based on the national electric vehicle ownership and average daily power consumption per vehicle, the national daily power demand is calculated, and the daily power demand for each station is derived. Waiting time and probability thresholds are set, the minimum station cost is calculated, and recommended station types and capacities are output. A candidate station set is constructed, and a comprehensive score is calculated. New stations are selected based on the comprehensive score to build a new station set. The annual and station-by-station power demand, station type, capacity, and new construction plans are integrated to generate an annual construction suggestion list and multi-scenario comparison results for rolling decision-making and scheme evaluation. This invention improves the stability, comparability, and applicability of annual station evaluation results.

Description

Technical Field

[0001] This invention specifically relates to a method for site selection and capacity determination of charging and battery swapping stations based on entropy weight method and queuing theory. Background Technology

[0002] With the continuous increase in the number of electric vehicles, charging and battery swapping infrastructure has entered a large-scale construction phase, and the focus of site planning has gradually shifted from simply increasing quantity to matching supply and demand. Significant differences exist across regions in terms of travel intensity, employment concentration, road network conditions, and the built environment, resulting in an uneven spatial distribution of energy replenishment demand. Simultaneously, site load fluctuates on both daily and interannual scales: insufficient capacity can lead to increased queuing times and decreased service levels; excessive capacity, on the other hand, can result in low equipment utilization and redundant investment. Therefore, it is necessary to calculate site demand over multiple years and under various scenarios, and further combine this with service capacity analysis to conduct capacity allocation, providing a basis for the rolling construction and optimization of the charging and battery swapping network.

[0003] In existing research and engineering practice, site demand assessment is mostly based on administrative region statistics, empirical coefficients, or multi-indicator evaluation results to determine the strength of demand at the regional level or the priority of site layout. Capacity allocation is often estimated using methods such as peak factor, empirical utilization rate, or pile-to-vehicle ratio. Although some studies have introduced queuing theory to analyze the service capacity of a single site, the relevant parameters are often directly given externally, lacking linkage with the results of large-scale demand assessment, and rarely considering the spatial differences and year-on-year changes between different sites. Therefore, existing methods usually struggle to form a continuous analytical framework from year-on-year demand assessment and service capacity verification to capacity allocation and new site layout. Summary of the Invention

[0004] Purpose of the invention: To provide a site selection and capacity determination method for charging and battery swapping stations based on entropy weight method and queuing theory, which solves the above-mentioned problems existing in the prior art.

[0005] Technical solution: A site selection and capacity determination method for charging and battery swapping stations based on entropy weight method and queuing theory, including the following steps:

[0006] S1. Acquire multi-source spatial data, and uniformly process the multi-source spatial data for coordinates and spatial reference to build an object dataset, which includes at least an existing site set. Parking lot candidate point set Year Collection Scenario Collection Build a candidate indicator set ;

[0007] S2, For the candidate index set Redundancy detection is performed on the candidate indicators to obtain the retained indicator set. ;

[0008] S3, For the retained index set The site in and indicators After defining the upper and lower quantiles, the positive and negative indices are calculated based on the upper and lower quantiles, and the corresponding years are calculated using the positive and negative indices. The proportional coefficient is calculated based on the proportional coefficient. After the proportional coefficient is completed, the information entropy of the year and the indicator is calculated based on the proportional coefficient. Finally, the year-by-year entropy weight is calculated. The time smoothing weight is preset and the smoothed weight is calculated. The year-by-year site score is obtained based on the smoothed weight.

[0009] S4. Using the scenario set as an index, calculate the national daily electricity demand by predicting the national electric vehicle ownership and the average daily electricity consumption per vehicle, and calculate the daily electricity demand of each station by conserving the distribution according to the annual station score.

[0010] S5. Obtain the site type and build a site type set, calculate the single service time of the sites in the site type set, obtain the daily electricity demand of the site and the single service time to calculate the vehicle arrival rate, preset the waiting time threshold and the waiting probability threshold, obtain the site with the minimum number of service stations based on the waiting time threshold and the waiting probability threshold to obtain the minimum service station, obtain the fixed construction cost and the unit service station cost, calculate the cost of the minimum service station to obtain the minimum station cost, and output the recommendation type and recommendation capacity.

[0011] S6. When the maximum number of feasible service units at an existing site under constraints of land use, power grid access, or investment threshold engineering is less than the recommended capacity output in step S5, the existing site is deemed to have insufficient capacity. Using the existing site as the center, parking lot POIs are retrieved within a preset search radius to construct a candidate site set. For each candidate site in the candidate site set, the quantile robust standardization and time-smoothing entropy weighting rules from step S3 are reused to calculate the comprehensive score of the candidate point. Based on the comprehensive score of the candidate point, a distance attenuation coefficient for "entry diversion" is introduced, and the comprehensive score of the candidate point is correlated with the distance from the candidate point to the existing site. The candidate sites are combined to obtain a comprehensive priority, wherein the comprehensive priority is obtained by weighting the comprehensive score by distance decay. The top F candidate sites are selected from high to low based on the comprehensive priority as the locations of new sites, thus constructing a set of new sites. The daily service demand of the existing sites under the feasible capacity is calculated, and the excess demand is obtained. The excess demand is then conserved and allocated according to the comprehensive priority of the new sites to obtain the daily demand of each new site. Using the daily demand of the new sites as input, step S5 is executed to output the recommended type and recommended capacity of each new site.

[0012] S7. Obtain the annual and station-by-station power demand, recommendation type, recommendation capacity, new site set, recommendation type and recommendation capacity of new sites to build an annual construction suggestion list and multi-scenario comparison results for rolling decision-making and scheme evaluation.

[0013] Preferably, the multi-source spatial data in step S1 includes at least population raster, nighttime light raster, road vector, road network density raster, charging and swapping stations, parking lots, sDNA indicators, building attributes, land use types, and grid access conditions for each existing station in the existing station cluster. Set up the site statistics window For each parking candidate point in the parking candidate point set Set candidate statistics window Site statistics window station radius and candidate statistics window candidate radius Define the scope of the analysis.

[0014] Preferably, the process of constructing the candidate indicator set in step S1 is as follows:

[0015] For each site ,years and indicators Calculate the original values ​​of candidate indicators within the scope of analysis. The candidate indicators include at least population density, road network density, nighttime light intensity, and employment attractiveness. Population density is calculated by dividing the total population within the analysis range by the area of ​​the analysis window. Road network density is calculated by dividing the total length of roads within the analysis range by the area of ​​the analysis range. Nighttime light intensity is calculated by calculating the average pixel value or area-weighted average value within the analysis range. Employment attractiveness is calculated by dividing the total building area of ​​industrial and commercial buildings by the area of ​​the analysis range.

[0016] Preferably, the calculation steps for retaining the index set in step S2 are as follows:

[0017] S2.1 Obtaining the candidate indicator set Any two indicators Calculate the rank correlation coefficient ;

[0018] S2.2, Preset configuration threshold When the rank correlation coefficient The absolute value is greater than or equal to the configured threshold. When calculating the distance correlation coefficient Mutual information Through distance correlation coefficient Mutual information To identify nonlinear dependencies and information redundancy;

[0019] S2.3. Based on the improvement in goodness of fit, correlation coefficient, or validation error of the target requirement variables, a retention decision is made to obtain the set of retention indicators. .

[0020] Preferably, the specific scoring steps for the annual site scoring in step S3 are as follows:

[0021] For each year Each indicator is subjected to quantile-robust standardization, and the retained indicator set is calculated using the following formula. The site in and indicators Define upper and lower quantiles:

[0022] In the formula: Indicates the year ,index The lower quantile; Indicates the year ,index The upper quantile; represent Quantile operator; Represents the quantile parameter. ; Indicates the original value of the candidate indicator;

[0023] Preset constants The formula for calculating the constant is as follows: In the formula: Indicates the number of existing sites; Represent the natural logarithm; then calculate the year using a preset constant. ,index Information entropy The information entropy is substituted into the following formula to calculate the yearly entropy weight:

[0024] In the formula: Indicates the year Entropy weight; Indicates the index variable; Indicates to the first Add the information entropy of each indicator to the numerical stability term. Among them, the numerical stability term The values ​​are positive to avoid the denominator being 0 or too small during normalization calculations; the time-smoothed weights are constructed using entropy weights. ;

[0025] Smoothed weights Substituting into the following formula, the annual site rating is calculated: In the formula: This indicates the site ratings over the years. Indicates whether it is a positive or negative indicator. This represents the smoothing weight.

[0026] Preferably, the calculation steps for the daily electricity demand of the station in step S4 are as follows:

[0027] With scenario collection The scene For indexing, the scenario ,years Input the national electric vehicle ownership forecast Average daily power consumption per bicycle To obtain the national daily electricity demand The daily electricity demand across the country is conservatively allocated based on the station rating, resulting in the station rating. The daily electricity demand, of which, the station The formula for calculating daily electricity demand is as follows:

[0028] In the formula: Indicates site In the year ,scene Daily electricity demand; Rate the site; Represents the sum of all site ratings; its distribution satisfies Conservation, error is The introduced numerical stability term leads to, This represents a numerically stable term.

[0029] Preferably, the specific steps for recommending type and recommending capacity in step S5 are as follows:

[0030] The site type set includes at least AC slow charging, DC fast charging, high-power DC supercharging, and battery swapping. The single service time for each charging and battery swapping site type in the set is calculated. The formula for calculating the single service time for charging sites is as follows:

[0031] In the formula: Indicates the single service time of a charging station; This indicates the average energy replenishment per cycle; Indicates rated power; Indicates the efficiency coefficient; the single service time for battery swapping sites is directly given. The service rate of the unit's service desk , Then through the site The daily electricity demand is converted into the vehicle arrival rate at this station. The calculation formula is as follows:

[0032] In the formula: Indicates site In the year ,scene Daily electricity demand; This indicates the average energy replenishment per cycle; This indicates the effective service duration per day.

[0033] Preferably, obtain the number of service stations configured at the site. The traffic intensity of the station is calculated using the following formula:

[0034] In the formula: Indicates traffic intensity, where the stability condition is: ; Indicates arrival rate; Indicates service rate; Indicates the number of service units;

[0035] The business volume parameter is calculated using the following formula: In the formula: This represents the traffic volume parameter. Based on the traffic intensity and traffic volume parameters, the system idle probability, waiting probability, and average waiting time are calculated. The formula for calculating the idle probability is as follows:

[0036] In the formula: Indicates on the site ,years ,scene Site type Below, the probability of the service console being idle within the system; Indicates the summation index; To represent factorial; Indicates traffic intensity. It is used to characterize the relative magnitude of arriving load and service capacity; This represents the business volume parameter. , express of Power of; Indicates the number of service counters; express factorial; express of Power; This indicates that the expression within the parentheses is taken as its reciprocal;

[0037] The formula for calculating the probability of waiting is as follows:

[0038] In the formula: express of Power; Indicates the probability that the system is idle; express factorial; Indicates traffic intensity; This indicates the probability that the vehicle will need to wait to arrive.

[0039] The formula for calculating the average waiting time is as follows:

[0040] In the formula: This indicates the average waiting time for vehicles in queues; This indicates the probability that the vehicle will need to wait to arrive. Indicates the number of service counters; Indicates the service rate of the unit's service desk; Indicates the vehicle arrival rate at the station; preset waiting time threshold. and waiting probability threshold Average vehicle waiting time Less than or equal to the waiting time threshold The probability that the vehicle will need to wait to arrive. Less than or equal to the waiting probability threshold Obtain the maximum capacity of the power grid connection for the site. Among them, the upper limit of capacitance Greater than or equal to total power demand The system employs an incremental search on the multi-source spatial data of each site. The minimum feasible capacity that satisfies the following constraints: , , , And mark the minimum feasible number of service units as Minimum number of feasible service units Indicates site In the year ,scene Charging station type Find the minimum number of service stations that satisfy the constraints, and then calculate the charging type of different stations in parallel for the same station. Minimum number of feasible service units The site with the lowest investment cost is selected as the recommended site. The formula for calculating the investment cost is as follows:

[0041] In the formula: Indicates investment cost; Indicates the charging type at the site. Fixed costs; The cost per service counter is used to calculate the referral type based on investment cost. The formula for calculating the referral type is as follows:

[0042] In the formula: Indicates the recommendation type; Indicates investment cost; then, by recommending types Substituting the values ​​into the following formula, you can obtain the recommended capacity. The formula is as follows:

[0043] In the formula: Indicates on the site ,years ,scene Below, when the site type is selected as the recommendation type. At that time, the corresponding minimum number of feasible service units; Indicates the recommendation type.

[0044] Preferably, the process of obtaining the recommendation type and recommendation capacity of the newly created site in step S6 is as follows:

[0045] When the expansion of existing sites is less than the recommended capacity, i.e., the maximum number of server desks implemented. If the existing site's capacity is insufficient, it is considered as... ; and search for parking lots in its domain to build a candidate set. Among them, candidate set , This represents the radius of the candidate search. Represents the set of parking lot POIs. Indicates candidate sites spatial coordinate vector, Indicates existing sites spatial coordinate vector, Indicate candidate sites, obtain candidate set Candidate sites Calculate the candidate site according to steps 3 and 4. Standardized indicators and smoothing weights The candidate point score is calculated using the following formula:

[0046] In the formula: Indicates candidate sites In the year The candidate site scores are then used to define the overall priority of the candidate sites, calculated using the following formula:

[0047] In the formula: Indicates candidate sites Compared to existing sites In the year The overall priority is as follows; Indicates candidate sites In the year Candidate point scores; Represents the distance decay function; Indicates existing sites With candidate sites The spatial distance between them; Get the highest priority Select candidate sites as locations for new sites to build a new site set. New site set ,in, Indicates the higher priority. One candidate site;

[0048] Calculate existing sites The daily electricity demand that can be served under the implemented capacity is calculated using the following formula:

[0049] In the formula: Indicates site Available daily electricity demand; Indicates a single energy replenishment; Indicates the daily effective service duration; Indicates the maximum number of service units that can be implemented; This indicates the unit service rate under the recommendation type.

[0050] Preferably, after calculating the daily electricity demand that can be served, the existing sites are then... The unmet daily electricity demand is calculated, and the existing substations that fail to meet daily electricity demand are included. After the power is allocated to the selected new site, the daily power demand to be undertaken by the new site is calculated. The daily power demand to be undertaken is then replaced with the daily power demand in step 5. Step 5 is then executed to obtain the recommended type and recommended capacity of the new site.

[0051] Beneficial effects: This invention relates to a method for site selection and capacity determination of charging and battery swapping stations based on entropy weight method and queuing theory, by analyzing a set of candidate indicators. Redundancy detection is performed on candidate indicators, and positive and negative indicators are calculated by combining upper and lower quantiles. The annual entropy weight and smoothed weight are improved, which can reduce the impact of outliers and interannual fluctuations on indicator weights and scoring results, thereby improving the stability, comparability and applicability of annual site evaluation results.

[0052] Establish a conservation distribution relationship between site rating results and total scenario demand, so that the evaluation results, which were originally mainly used for relative ranking, can be further transformed into the annual electricity demand of the site, providing a unified demand input for subsequent capacity configuration;

[0053] By combining the recommended access types and capacity constraints based on daily electricity demand, the required number of charging piles or battery swapping service capacity for each site is back-calculated and verified, transforming capacity configuration from empirical estimation to quantitative configuration based on demand levels and service constraints. Furthermore, this method can adapt to differences in service time and parallel service capabilities among different charging methods.

[0054] When existing sites cannot meet the demand due to limitations in capacity, land use, or access conditions, candidate new sites are generated through neighboring parking lots, and corresponding capacity configuration results are provided simultaneously, thereby forming a construction plan that connects expansion and new construction, and improving the feasibility of the results.

[0055] It obtains the annual and station-by-station power demand, recommended type, recommended capacity, new site set, recommended type and recommended capacity of new sites, and outputs the annual site demand, capacity configuration results and construction suggestion list, providing reusable technical basis for the rolling construction, scheme evaluation and investment decision of charging and swapping networks. Attached Figure Description

[0056] Figure 1 This is a system block diagram of the present invention;

[0057] Figure 2 This represents the demand and recommended configuration results for representative sites of this invention. Detailed Implementation

[0058] like Figures 1 to 2 As shown, this invention provides a technical solution: a method for site selection and capacity determination of charging and battery swapping stations based on entropy weight method and queuing theory, comprising the following steps:

[0059] S1. Acquire multi-source spatial data, which includes at least population raster, nighttime light raster, road vector, road network density raster, charging and swapping stations, parking lots, sDNA indicators, building attributes, land use types, and power grid access conditions for each existing station in the existing station cluster. Set up the site statistics window For the candidate point set of parking lots Each parking lot candidate point Set candidate statistics window Site statistics window station radius and candidate statistics window candidate radius Define the scope of analysis, and uniformly process the coordinates and spatial references of the multi-source spatial data to build an object dataset, which includes at least the existing site set. Parking lot candidate point set Year Collection Scenario Collection Build a candidate indicator set The process of building the candidate indicator set is as follows:

[0060] For each site ,years and indicators Calculate the original values ​​of candidate indicators within the scope of analysis. The candidate indicators include at least population density, road network density, nighttime light intensity, and employment attractiveness. Population density is calculated by dividing the total population within the analysis range by the area of ​​the analysis window. Road network density is calculated by dividing the total length of roads within the analysis range by the area of ​​the analysis range. Nighttime light intensity is calculated by calculating the average pixel value or area-weighted average value within the analysis range. Employment attractiveness is calculated by dividing the total building area of ​​industrial and commercial buildings by the area of ​​the analysis range.

[0061] S2, for each station ,years and indicators Calculate the original values ​​of candidate indicators within the statistical window. The calculation formula is as follows:

[0062] In the formula, Indicates the year The Data source class; This represents the window statistics sub-calculator corresponding to this indicator; This represents the site statistics window; for the candidate indicator set Redundancy detection is performed on the candidate indicators to obtain the retained indicator set. The calculation steps for retaining the index set are as follows:

[0063] S2.1 Obtaining the candidate indicator set Any two indicators Calculate the rank correlation coefficient , where the rank correlation coefficient The calculation formula is as follows:

[0064] In the formula: Indicators and rank correlation coefficient; Indicates the first The rank difference of an observation between two variables; Indicates the size of the observed sample;

[0065] S2.2, Preset configuration threshold When the rank correlation coefficient The absolute value is greater than or equal to the configured threshold. When calculating the distance correlation coefficient Mutual information Through distance correlation coefficient Mutual information To identify nonlinear dependencies and information redundancy;

[0066] Among them, the distance correlation coefficient The calculation formula is as follows:

[0067] In the formula, Represents the distance correlation coefficient; This represents the distance-related calculation operator; Indicators The corresponding candidate indicator's original value; Indicators The corresponding candidate indicator's original value;

[0068] Mutual Information The calculation formula is as follows:

[0069] In the formula, Represents mutual information; Represents the mutual information computation operator;

[0070] S2.3. Based on the improvement in goodness of fit, correlation coefficient, or validation error of the target requirement variables, a retention decision is made to obtain the set of retention indicators. .

[0071] S3, For the retained index set The site in and indicators After defining the upper and lower quantiles, the positive and negative indices are calculated based on the upper and lower quantiles, and the corresponding years are calculated using the positive and negative indices. The proportional coefficient is used to calculate the information entropy of the year and the indicator. Finally, the yearly entropy weight is calculated. A time smoothing weight is preset and calculated. Based on the smoothed weight, the yearly site score is obtained. The specific scoring steps for the yearly site score are as follows:

[0072] For each year Each indicator is subjected to quantile-robust standardization, and the retained indicator set is calculated using the following formula. The site in and indicators Define upper and lower quantiles:

[0073] In the formula: Indicates the year ,index The lower quantile; Indicates the year ,index The upper quantile; represent Quantile operator; Represents the quantile parameter. ; Indicates the original value of the candidate indicator;

[0074] The formulas for calculating the positive and negative indices based on the upper and lower quantiles are as follows:

[0075] ; In the formula: This represents the standardized index value; Indicates will Cut off to ; This indicates the smallest positive number where the denominator is 0; the following formula is used to calculate the value in the year. The following proportionality coefficient:

[0076] In the formula: Indicates site In the year ,index The proportion; This represents the sum of the standardized values ​​of all sites for this metric; a preset constant. The formula for calculating the constant is as follows: In the formula: Indicates the number of existing sites; Represent the natural logarithm; then calculate the year using a preset constant. ,index Information entropy The information entropy is substituted into the following formula to calculate the yearly entropy weight:

[0077] In the formula: Indicates the year Entropy weight; Indicates the index variable; Indicates to the first Add the information entropy of each indicator to the numerical stability term. Among them, the numerical stability term The values ​​are positive to avoid the denominator being 0 or too small during normalization calculations; the time-smoothed weights are constructed using entropy weights. Among them, the weights after time smoothing The calculation formula is as follows:

[0078] In the formula: Indicates the smoothed weights; This represents the smoothing coefficient. In this embodiment, the smoothing coefficient is 0.5, which is configurable. This represents the smoothing weight of the previous year; the initial value can be... ; smoothed weights Substituting into the following formula, the annual site rating is calculated: In the formula: This indicates the site ratings over the years. Indicates whether it is a positive or negative indicator. This represents the smoothing weight.

[0079] S4, using scenario sets The scene For indexing, the scenario ,years Input the national electric vehicle ownership forecast Average daily power consumption per bicycle The formula for calculating the national daily electricity demand is as follows:

[0080] In the formula: Indicates the year ,scene Daily electricity demand across the country; This indicates a forecast of the total number of electric vehicles in China. This represents the average daily power consumption per vehicle; and the daily power demand of each station is calculated using a conservation distribution based on the annual station rating. The calculation steps for the daily power demand of each station are as follows:

[0081] The daily electricity demand across the country is conservatively allocated based on the station rating, resulting in the station rating. The daily electricity demand, of which, the station The formula for calculating daily electricity demand is as follows:

[0082] In the formula: Indicates site In the year ,scene Daily electricity demand; Rate the site; Represents the sum of all site ratings; its distribution satisfies Conservation, error is The introduced numerical stability term leads to, This represents a numerically stable term.

[0083] S5. Obtain Site Types and Build a Site Type Set Among them, the site type set At least include AC slow charging type DC fast charging type Supercharging type and battery swapping type For each type Define its key service parameters, including rated power. Applicable to charging types, including AC slow charging. DC fast charging type and supercharging type Single service time Suitable for battery swapping type Or, for charging types used to directly specify service time; average energy replenishment per charge. This indicates the expected energy replenishment value for a single vehicle refueling operation under this type of service; effective service duration. Indicates site Effective daily service availability; charging efficiency This method incorporates efficiency factors in power and energy conversion into calculations; it calculates the single service time for stations in a station type set, obtains the daily electricity demand and single service time of each station to calculate the vehicle arrival rate, presets waiting time and waiting probability thresholds, and obtains the minimum number of service stations based on these thresholds to determine the minimum service station. It also obtains the fixed construction cost and unit service station cost, calculates the cost of the minimum service station to determine the minimum station cost, and outputs recommended types and recommended capacity. The specific steps for recommending types and recommended capacity are as follows:

[0084] The site type set includes at least AC slow charging, DC fast charging, high-power DC supercharging, and battery swapping. The single service time for each charging and battery swapping site type in the set is calculated. The formula for calculating the single service time for charging sites is as follows:

[0085] In the formula: Indicates the single service time of a charging station; This indicates the average energy replenishment per cycle; Indicates rated power; Indicates the efficiency coefficient; the single service time for battery swapping sites is directly given. The service rate of the unit's service desk , Then through the site The daily electricity demand is converted into the vehicle arrival rate at this station. The calculation formula is as follows:

[0086] In the formula: Indicates site In the year ,scene Daily electricity demand; This indicates the average energy replenishment per cycle; This indicates the effective service duration per day.

[0087] Get the number of service nodes configured on the site The traffic intensity of the station is calculated using the following formula:

[0088] In the formula: Indicates traffic intensity, where the stability condition is: ; Indicates arrival rate; Indicates service rate; Indicates the number of service units;

[0089] The business volume parameter is calculated using the following formula: In the formula: This represents the traffic volume parameter. Based on the traffic intensity and traffic volume parameters, the system idle probability, waiting probability, and average waiting time are calculated. The formula for calculating the idle probability is as follows:

[0090] In the formula: Indicates on the site ,years ,scene Site type Below, the probability of the service console being idle within the system; Indicates the summation index; To represent factorial; Indicates traffic intensity. It is used to characterize the relative magnitude of arriving load and service capacity; This represents the business volume parameter. , express of Power; Indicates the number of service counters; express factorial; express of Power; This indicates that the expression within the parentheses is taken as its reciprocal;

[0091] The formula for calculating the probability of waiting is as follows:

[0092] In the formula: express of Power; Indicates the probability that the system is idle; express factorial; Indicates traffic intensity; This indicates the probability that the vehicle will need to wait to arrive.

[0093] The formula for calculating the average waiting time is as follows:

[0094] In the formula: This indicates the average waiting time for vehicles in queues; This indicates the probability that the vehicle will need to wait to arrive. Indicates the number of service counters; Indicates the service rate of the unit's service desk; Indicates the vehicle arrival rate at the station; preset waiting time threshold. and waiting probability threshold Set service level thresholds , Average vehicle waiting time Less than or equal to the waiting time threshold The probability that the vehicle will need to wait to arrive. Less than or equal to the waiting probability threshold Obtain the maximum capacity of the power grid connection for the site. Among them, the upper limit of capacitance Greater than or equal to total power demand In a further embodiment, a land use / workstation cap is introduced. ,Require: ,in, Indicates site The system employs an incremental search for multi-source spatial data at each site, based on the maximum number of service stations allowed by site constraints. The minimum feasible capacity that satisfies the following constraints: , , , And mark the minimum feasible number of service units as Minimum number of feasible service units Indicates site In the year ,scene Charging station type Find the minimum number of service stations that satisfy the constraints, and then calculate the charging type of different stations in parallel for the same station. Minimum number of feasible service units The site with the lowest investment cost is selected as the recommended site. The formula for calculating the investment cost is as follows:

[0095] In the formula: Indicates investment cost; Indicates the charging type at the site. Fixed costs; The cost per service counter is used to calculate the referral type based on investment cost. The formula for calculating the referral type is as follows:

[0096] In the formula: Indicates the recommendation type; Indicates investment cost; then, by recommending types Substituting the values ​​into the following formula, you can obtain the recommended capacity. The formula is as follows:

[0097] In the formula: Indicates on the site ,years ,scene Below, when the site type is selected as the recommendation type. At that time, the corresponding minimum number of feasible service units; Indicates the recommendation type.

[0098] S6. When the maximum number of feasible service units at an existing site under constraints of land use, power grid access, or investment threshold engineering is less than the recommended capacity output in step S5, the existing site is deemed to have insufficient capacity. Using the existing site as the center, parking lot POIs are retrieved within a preset search radius to construct a candidate site set. For each candidate site in the candidate site set, the quantile robust standardization and time-smoothing entropy weighting rules from step S3 are reused to calculate the comprehensive score of the candidate point. Based on the comprehensive score of the candidate point, a distance attenuation coefficient for "entry diversion" is introduced, and the comprehensive score of the candidate point is correlated with the distance from the candidate point to the existing site. The candidate sites are combined to obtain a comprehensive priority, wherein the comprehensive priority is obtained by weighting the comprehensive score by distance decay. The top F candidate sites are selected from high to low based on the comprehensive priority as the locations of new sites, thus constructing a set of new sites. The daily service demand of the existing sites under the feasible capacity is calculated, and the excess demand is obtained. The excess demand is then conserved and allocated according to the comprehensive priority of the new sites to obtain the daily demand of each new site. Using the daily demand of the new sites as input, step S5 is executed to output the recommended type and recommended capacity of each new site.

[0099] The process of determining the recommendation type and recommendation capacity for newly created sites is as follows:

[0100] When the expansion of existing sites is less than the recommended capacity, i.e., the maximum number of server desks implemented. If the existing site's capacity is insufficient, it is considered as... ; and search for parking lots in its domain to build a candidate set. Among them, candidate set , This represents the radius of the candidate search. Represents the set of parking lot POIs. Indicates candidate sites spatial coordinate vector, Indicates existing sites spatial coordinate vector, Indicate candidate sites, obtain candidate set Candidate sites Calculate the candidate site according to steps 3 and 4. Standardized indicators and smoothing weights The candidate point score is calculated using the following formula:

[0101] In the formula: Indicates candidate sites In the year The candidate site scores are then used to define the overall priority of the candidate sites, calculated using the following formula:

[0102] In the formula: Indicates candidate sites Compared to existing sites In the year The overall priority is as follows; Indicates candidate sites In the year Candidate point scores; Represents the distance decay function; Indicates existing sites With candidate sites In this embodiment, the distance attenuation function is selected to represent the spatial distance between them. , , ;in Get the highest priority Select candidate sites as locations for new sites to build a new site set. New site set ,in, Indicates the higher priority. One candidate site;

[0103] Calculate existing sites The daily electricity demand that can be served under the implemented capacity is calculated using the following formula:

[0104] In the formula: Indicates site Available daily electricity demand; Indicates a single energy replenishment; Indicates the daily effective service duration; Indicates the maximum number of service units that can be implemented; This indicates the unit service rate under the recommended type. After calculating the daily available electricity demand, the existing sites are then... The unmet daily electricity demand is calculated, and the existing substations that fail to meet daily electricity demand are included. After the power is allocated to the selected new site, the daily power demand to be undertaken by the new site is calculated. The daily power demand to be undertaken is then replaced with the daily power demand in step 5. Step 5 is then executed to obtain the recommended type and recommended capacity of the new site.

[0105] S7. Obtain the annual and station-by-station power demand, recommendation type, recommendation capacity, new site set, recommendation type and recommendation capacity of new sites to build an annual construction suggestion list and multi-scenario comparison results for rolling decision-making and scheme evaluation.

[0106] The following example, using a public charging scenario in Jiangsu Province, illustrates the effectiveness of this invention in solving the problems of existing technologies. After redundancy detection, this embodiment uses four indicators—population density, road network density, nighttime light intensity, and employment attractiveness—and obtains the indicator weights using the entropy weight method. In the demand pushdown stage, demand is allocated based on the site scoring conservation principle. In the capacity determination stage, the minimum feasible charging gun scale is calculated backwards.

[0107] like Figure 2 As shown, the site rating results, which were originally only used for relative ranking, are further transformed into site demand, and the site capacity configuration results are output under the same analysis framework. At the same time, it can identify high-demand core areas and traffic corridor nodes, and control the average waiting time within a preset threshold, thus forming a continuous decision-making chain from demand measurement to capacity configuration and then to the generation of new sites.

[0108] As shown in Table 1, in this embodiment, the weights of employment attractiveness and road network density are significantly higher than those of population density and nighttime light intensity, indicating that the present invention can objectively identify the dominant factors affecting the priority of site layout based on the degree of data dispersion.

[0109] Table 1. Entropy weighting results in the provincial implementation example.

[0110] Serial Number index Weight 1 population density 0.172812 2 Road network density 0.286140 3 Nighttime light intensity 0.016273 4 Job attraction 0.524775

[0111] like Figure 2 As shown, this invention not only provides demand capacity results for representative sites, but also derives the corresponding minimum feasible configuration size under a given waiting time threshold. The average waiting time for the aforementioned representative sites does not exceed 8 minutes, indicating that this invention can directly transform service level constraints into capacity configuration results, avoiding reliance solely on the pile-to-vehicle ratio or empirical utilization rate for capacity determination.

[0112] Therefore, the solution of this invention can simultaneously solve the problems in the prior art such as "the scoring results are difficult to convert into site demand input", "the arrival rate parameters lack a unified source", "the capacity configuration lacks service level constraints" and "the lack of linkage mechanism for adding new sites when existing sites are insufficient", and form an executable expansion and new construction suggestion result.

[0113] The preferred embodiments of the present invention have been described in detail above. However, the present invention is not limited to the specific details in the above embodiments. Within the scope of the technical concept of the present invention, various equivalent transformations can be made to the technical solutions of the present invention, and these equivalent transformations all fall within the protection scope of the present invention.

Claims

1. A method for site selection and capacity determination of charging and battery swapping stations based on entropy weight method and queuing theory, characterized in that, Includes the following steps: S1. Acquire multi-source spatial data, and uniformly process the multi-source spatial data for coordinates and spatial reference to build an object dataset, which includes at least an existing site set. Parking lot candidate point set Year Collection Scenario Collection Build a candidate indicator set ; S2, For the candidate index set Redundancy detection is performed on the candidate indicators to obtain the retained indicator set. ; S3, For the retained index set The site in and indicators After defining the upper and lower quantiles, the positive and negative indices are calculated based on the upper and lower quantiles, and then the corresponding year is calculated. The proportional coefficient and annual entropy weight; preset time smoothing weight and calculate the smoothed weight, and obtain the annual site score based on the smoothed weight; S4. Using the scenario set as an index, calculate the national daily electricity demand by predicting the national electric vehicle ownership and the average daily electricity consumption per vehicle, and calculate the daily electricity demand of each station by conserving the distribution according to the annual station score. S5. Obtain the site type and build a site type set, calculate the single service time of the sites in the site type set, obtain the daily electricity demand of the site and the single service time to calculate the vehicle arrival rate, preset the waiting time threshold and the waiting probability threshold, obtain the site with the minimum number of service stations based on the waiting time threshold and the waiting probability threshold to obtain the minimum service station, obtain the fixed construction cost and the unit service station cost, calculate the cost of the minimum service station to obtain the minimum station cost, and output the recommendation type and recommendation capacity. S6. When the maximum number of service stations that can be implemented at an existing site under the constraints of land use, power grid access, or investment threshold engineering is less than the recommended capacity output in step S5, the existing site is determined to have insufficient capacity. Centered on the existing site, parking lot POIs are retrieved within a preset search radius to construct a candidate site set. For each candidate site in the candidate site set, the quantile robust standardization and time-smoothing entropy weighting rules of step S3 are reused to calculate the comprehensive score of the candidate point. Based on the comprehensive score of the candidate point, a distance attenuation coefficient for "entry diversion" is introduced, and the comprehensive score of the candidate point is combined with the distance from the candidate point to the existing site to obtain the comprehensive priority of the candidate site. The comprehensive priority is obtained by weighting the comprehensive score by distance attenuation. The top 5 candidate sites are selected as the locations of new sites according to the comprehensive priority from high to low to construct a new site set. The daily service demand of the existing site under the feasible capacity is calculated and the excess demand is obtained. The excess demand is conserved and allocated according to the comprehensive priority of the new sites to obtain the daily demand of each new site. With the daily demand of the new sites as input, step S5 is executed to output the recommended type and recommended capacity of each new site. S7. Obtain the annual and station-by-station power demand, recommendation type, recommendation capacity, new site set, recommendation type and recommendation capacity of new sites, build an annual construction suggestion list and multi-scenario comparison results for rolling decision-making and scheme evaluation.

2. The method for site selection and capacity determination of charging and battery swapping stations based on entropy weight method and queuing theory according to claim 1, characterized in that, The multi-source spatial data mentioned in step S1 includes at least population raster, nighttime light raster, road vector, road network density raster, charging and swapping stations, parking lots, sDNA indicators, building attributes, land use types, and grid access conditions for each existing station in the existing station cluster. Set up the site statistics window For each parking candidate point in the parking candidate point set Set candidate statistics window Site statistics window station radius and candidate statistics window candidate radius Define the scope of the analysis.

3. The method for site selection and capacity determination of charging and battery swapping stations based on entropy weight method and queuing theory according to claim 2, characterized in that, The process of building the candidate indicator set in step S1 is as follows: For each site ,years and indicators Calculate the original values ​​of candidate indicators within the scope of analysis. The candidate indicators include at least population density, road network density, nighttime light intensity, and employment attractiveness. Population density is calculated by dividing the total population within the analysis range by the area of ​​the analysis window. Road network density is calculated by dividing the total length of roads within the analysis range by the area of ​​the analysis range. Nighttime light intensity is calculated by calculating the average pixel value or area-weighted average value within the analysis range. Employment attractiveness is calculated by dividing the total building area of ​​industrial and commercial buildings by the area of ​​the analysis range.

4. The method for site selection and capacity determination of charging and battery swapping stations based on entropy weight method and queuing theory according to claim 3, characterized in that, The calculation steps for retaining the index set in step S2 are as follows: S2.1 Obtaining the candidate indicator set Any two indicators Calculate the rank correlation coefficient ; S2.2, Preset configuration threshold When the rank correlation coefficient The absolute value is greater than or equal to the configured threshold. When calculating the distance correlation coefficient Mutual information Through distance correlation coefficient Mutual information To identify nonlinear dependencies and information redundancy; S2.

3. Based on the improvement in goodness of fit, correlation coefficient, or validation error of the target requirement variables, a retention decision is made to obtain the set of retention indicators. .

5. The method for site selection and capacity determination of charging and battery swapping stations based on entropy weight method and queuing theory according to claim 2, characterized in that, The specific scoring steps for the annual site rating in step S3 are as follows: For each year Each indicator is subjected to quantile-robust standardization, and the retained indicator set is calculated using the following formula. The site in and indicators Define upper and lower quantiles: In the formula: Indicates the year ,index The lower quantile; Indicates the year ,index The upper quantile; represent Quantile operator; Indicates the quantile parameter. ; This represents the original value of the candidate indicator; Preset constants The formula for calculating the constant is as follows: In the formula: Indicates the number of existing sites; Represent the natural logarithm; then calculate the year using a preset constant. ,index Information entropy The information entropy is substituted into the following formula to calculate the yearly entropy weight: In the formula: Indicates the year Entropy weight; Indicates the index variable; Indicates to the first The information entropy of each indicator is added to a numerical stability term. Among them, the numerical stability term The values ​​are positive to avoid the denominator being 0 or too small during normalization calculations; the time-smoothed weights are constructed using entropy weights. ; Smoothed weights Substituting into the following formula, the annual site rating is calculated: In the formula: This indicates the site ratings over the years. Indicates whether it is a positive or negative indicator. This represents the smoothing weight.

6. The method for site selection and capacity determination of charging and battery swapping stations based on entropy weight method and queuing theory according to claim 1, characterized in that, The calculation steps for the daily electricity demand of the station in step S4 are as follows: With scenario collection The scene For indexing, the scenario ,years Input the national electric vehicle ownership forecast Average daily power consumption per bicycle To obtain the national daily electricity demand The daily electricity demand across the country is conservatively allocated based on the station rating, resulting in the station rating. The daily electricity demand, of which, the station The formula for calculating daily electricity demand is as follows: In the formula: Indicates site In the year ,scene Daily electricity demand; Rate the site; Represents the sum of all site ratings; its distribution satisfies Conservation, error is The introduced numerical stability term leads to, This represents a numerically stable term.

7. The method for site selection and capacity determination of charging and battery swapping stations based on entropy weight method and queuing theory according to claim 6, characterized in that, The specific steps for recommending type and recommending capacity in step S5 are as follows: The site type set includes at least AC slow charging, DC fast charging, high-power DC supercharging, and battery swapping. The single service time for each charging and battery swapping site type in the set is calculated. The formula for calculating the single service time for charging sites is as follows: In the formula: Indicates the single service time of a charging station; This indicates the average energy replenishment per cycle; Indicates rated power; Indicates the efficiency coefficient; the single service time for battery swapping sites is directly given. The service rate of the unit's service desk , Then through the site The daily electricity demand is converted into the vehicle arrival rate at this station. The calculation formula is as follows: In the formula: Indicates site In the year ,scene Daily electricity demand; This indicates the average energy replenishment per cycle; This indicates the effective service duration per day.

8. The method for site selection and capacity determination of charging and battery swapping stations based on entropy weight method and queuing theory according to claim 7, characterized in that, Get the number of service nodes configured on the site The traffic intensity of this station is calculated using the following formula: In the formula: Indicates traffic intensity, where the stability condition is: ; Indicates arrival rate; Indicates service rate; Indicates the number of service units; The business volume parameter is calculated using the following formula: In the formula: This represents the traffic volume parameter. Based on the traffic intensity and traffic volume parameters, the system idle probability, waiting probability, and average waiting time are calculated. The formula for calculating the idle probability is as follows: In the formula: Indicates on the site ,years ,scene Site type Below, the probability of the service console being idle within the system; Indicates the summation index; To represent factorial; Indicates traffic intensity. It is used to characterize the relative magnitude of arriving load and service capacity; This represents the business volume parameter. , express of Power; Indicates the number of service counters; express factorial; express of Power; This indicates that the expression within the parentheses is taken as its reciprocal; The formula for calculating the probability of waiting is as follows: In the formula: express of Power; Indicates the probability that the system is idle; express factorial; Indicates traffic intensity; This indicates the probability that the vehicle will need to wait to arrive. The formula for calculating the average waiting time is as follows: In the formula: This indicates the average waiting time for vehicles in queues; This indicates the probability that the vehicle will need to wait to arrive. Indicates the number of service counters; Indicates the service rate of the unit's service desk; Indicates the vehicle arrival rate at the station; preset waiting time threshold. and waiting probability threshold Average vehicle waiting time Less than or equal to the waiting time threshold The probability that the vehicle will need to wait to arrive. Less than or equal to the waiting probability threshold Obtain the maximum capacity of the power grid connection for the site. Among them, the upper limit of the capacitor Greater than or equal to total power demand The system employs an incremental search on the multi-source spatial data of each site. The minimum feasible capacity that satisfies the following constraints: , , , And mark the minimum feasible number of service units as Minimum number of feasible service units Indicates site In the year ,scene Charging station type Find the minimum number of service stations that satisfy the constraints, and then calculate the charging type of different stations in parallel for the same station. Minimum number of feasible service units The site with the lowest investment cost is selected as the recommended site. The formula for calculating the investment cost is as follows: In the formula: Indicates investment cost; Indicates the charging type at the site Fixed costs; The cost per service counter is used to calculate the referral type based on investment cost. The formula for calculating the referral type is as follows: In the formula: Indicates the recommendation type; Indicates investment cost; then, by recommending types Substituting the values ​​into the following formula, you can obtain the recommended capacity. The formula is as follows: In the formula: Indicates on the site ,years ,scene Below, when the site type is selected as the recommendation type. At that time, the corresponding minimum number of feasible service units; Indicates the recommendation type.

9. The method for site selection and capacity determination of charging and battery swapping stations based on entropy weight method and queuing theory according to claim 8, characterized in that, The process of obtaining the recommendation type and recommendation capacity of the newly created site in step S6 is as follows: When the expansion of existing sites is less than the recommended capacity, i.e., the maximum number of server desks implemented. If the existing site's capacity is insufficient, it is considered as... ; and search for parking lots in its domain to build a candidate set. Among them, candidate set , This represents the radius of the candidate search. Represents the set of parking lot POIs. Indicates candidate sites spatial coordinate vector, Indicates existing sites spatial coordinate vector, Indicate candidate sites, obtain candidate set Candidate sites Calculate the candidate site according to steps 3 and 4. Standardized indicators and smoothing weights The candidate point score is calculated using the following formula: In the formula: Indicates candidate sites In the year The candidate site scores are then used to define the overall priority of the candidate sites, calculated using the following formula: In the formula: Indicates candidate sites Compared to existing sites In the year The overall priority is as follows; Indicates candidate sites In the year Candidate point scores; Represents the distance decay function; Indicates existing sites With candidate sites The spatial distance between them; Get the highest priority Select candidate sites as locations for new sites to build a new site set. New site set ,in, Indicates the higher priority. One candidate site; Calculate existing sites The daily electricity demand that can be served under the implemented capacity is calculated using the following formula: In the formula: Indicates site Available daily electricity demand; Indicates a single energy replenishment; Indicates the daily effective service duration; Indicates the maximum number of service units that can be implemented; This indicates the unit service rate under the recommendation type.

10. The method for site selection and capacity determination of charging and battery swapping stations based on entropy weight method and queuing theory according to claim 9, characterized in that, After calculating the daily electricity demand that can be served, then the existing sites are... The unmet daily electricity demand is calculated, and the existing substations that fail to meet daily electricity demand are included. After the power is allocated to the selected new site, the daily power demand to be undertaken by the new site is calculated. The daily power demand to be undertaken is then replaced with the daily power demand in step 5. Step 5 is then executed to obtain the recommended type and recommended capacity of the new site.

Images