Search

CN-121169470-B - Electric vehicle battery replacement station site selection optimization method considering multi-facility synchronous service

CN121169470BCN 121169470 BCN121169470 BCN 121169470BCN-121169470-B

Abstract

The invention discloses an electric automobile power exchange station site selection optimization method considering multi-facility synchronous service, which relates to the technical field of traffic operation management and comprises the following steps of 1, constructing a power exchange station site selection model, 2, constructing a queuing system and a multi-layer perceptron neural network structure, 3, generating a neighborhood by man-machine cooperation, executing a variable neighborhood search algorithm to solve the neighborhood, and obtaining an optimal power exchange station site selection scheme.

Inventors

  • LI XIANG
  • XU KANG
  • ZHANG BOWEN

Assignees

  • 长安大学

Dates

Publication Date
20260508
Application Date
20250902

Claims (6)

  1. 1. An electric vehicle battery replacement station site selection optimization method considering multi-facility synchronous service is characterized by comprising the following steps: Step 1, constructing a power exchange station site selection model; s11, determining the power change requirement, potential site selection points and road network information in a target area; S12, constructing a power exchange station site selection model, wherein the power exchange station site selection model aims at maximizing the battery replacement requirements of all power exchange stations, and takes constraint conditions that the average waiting time of drivers in a target area is not more than a preset threshold value and the utilization rate of each power exchange station is not less than the preset threshold value; step 2, constructing a queuing system and a multi-layer perceptron neural network structure; S21, regarding the power exchange station as a service facility, and under the condition that a plurality of service stations are arranged in the service facility, constructing the whole power exchange process into a queuing system based on M/G/K, wherein the expression of the queuing system is as follows: ...(16); ...(17); ...(18); ...(19); in the formula, Is shown in the first Standard deviation of time spent for battery replacement service of the battery replacement station at each position point; Represent the first Average waiting time of drivers whose service time of the power exchange stations at the individual location points obeys poisson distribution; Represent the first The service time of the power exchange station of each position point obeys the average waiting time of drivers in fixed distribution; Represent the first Average waiting time of drivers at the individual location points; Represent the first The total service rate of battery replacement service is carried out by the battery replacement stations at the plurality of position points; representing a set of alternative service types when constructing the power exchange station, ; For decision variables, represent the first Whether or not to build the first place A power exchange station of the seed service type takes 1 if the seed service type is the seed service type, and takes 0 if the seed service type is the seed service type; Represent the first Service capability of a service type of the power exchange station; Represent the first Digital features of battery replacement requirements faced by a battery replacement station at a plurality of locations; Represent the first The number of location point consoles; S22, constructing a multi-layer perceptron neural network structure, obtaining the average waiting time of the power exchange station according to the actual power exchange scene, and training the neural network structure; step 3, generating a neighborhood by man-machine cooperation, and executing a variable neighborhood searching algorithm to solve the neighborhood to obtain an optimal power exchange station site selection scheme; s31, embedding a neural network structure into a variable neighborhood search algorithm to obtain an estimated value of the waiting time of the power exchange station; s32, designing a prompt word interacted with the large language model; s33, designing a neighborhood structure; s34, generating a new neighborhood structure after interaction of the neighborhood structure and the large language model, and solving.
  2. 2. The method for optimizing the location of an electric vehicle battery exchange station taking into account multi-facility synchronous service according to claim 1, wherein the objective function of the battery exchange station location model consists of a calculation formula of the battery exchange requirement of the satisfied driver, and the specific expression is: ...(1); in the formula, Representing the set of location points in the city where the driver's battery replacement needs arise, ; Representing a collection Is selected from the group consisting of, Wherein A service radius that can be covered by the power exchange station; representing a set of potential location points in the city that meet the condition of construction of the power exchange station, ; Represent the first The position point is to the first The travel distance of the individual location points; Represent the first Digital features of battery replacement demand generated at the individual location points; for decision variables, represent the first Whether the station at the location is the first And the battery replacement demand service of each position point is 1 if the service is yes, and 0 if the service is not yes.
  3. 3. The electric vehicle battery exchange station site selection optimization method considering multi-facility synchronous service according to claim 2, wherein constraint functions of the battery exchange station site selection model are respectively as follows: the total investment of the construction scheme of the current power exchange station does not exceed budget, and the expression is as follows: ...(2); in the formula, Is shown in the first The fixed cost of the power exchange station is built on each position point; for decision variables, represent the first Whether a power exchange station is built on each position point or not, if yes, taking 1, and if not, taking 0; Is shown in the first Build on the site Varying costs of service desks of the generic service type; representing a budget for constructing the power exchange station; within the preset range of the constructed power exchange station, the power exchange requirement of the driver is served by the corresponding power exchange station, and the expression is as follows: ...(3); after the driver generates the power change requirement, selecting one from the power change stations in all coverage areas to go to and receive battery change service, wherein the expression is as follows: ...(4); the power exchange station constructed on one position point is provided with one service type of a plurality of service types, and the expression is as follows: ...(5); the total service capacity calculation expression of the power exchange station constructed on one position point is as follows: ...(6); in the formula, For decision variables, represent the first The number of the service desks built on each position point; the number of stations placed at a location is expressed as: ...(7); in the formula, Represent the first Maximum number of construction service desks on each position point; The expression of the battery replacement demand of the battery replacement station service driver at one position point is as follows: ...(8); When there are multiple stations within the coverage area that are acceptable for service, the driver goes to the nearest station, expressed as: ...(9); in the formula, Represent the first Station and a first station of a plurality of location points The distance between the individual demand points; Represent the first Whether or not to serve the first place If yes, taking 1, and if not, taking 0; the average waiting time of the driver after reaching the power exchange station is not higher than a threshold value The expression of (2) is: ...(10); in the formula, An upper threshold value representing the waiting time after the driver reaches the power exchange station; the service capacity of the power exchange station is not lower than the total demand of the service, and the expression is: ...(11); decision variables The value of (2) is 0 or 1, and the expression is: ...(12); decision variables The value of (2) is 0 or 1, and the expression is: ...(13); decision variables The value of (2) is 0 or 1, and the expression is: ...(14); decision variables The value of (2) is the interval The expression is: ...(15)。
  4. 4. The method for optimizing the site selection of the electric automobile power exchange station taking multi-facility synchronous service into consideration as claimed in claim 3, wherein the neural network structure in S22 comprises 3 hidden layers, the node numbers of the hidden layers are 64, 32 and 16 respectively, reLu is taken as an activation function, the learning rate is 5×10 -6 , adam is taken as an optimizer, and the loss function is MSE.
  5. 5. The method for optimizing the site selection of an electric vehicle battery exchange station taking into account multi-facility synchronous service as defined in claim 4, wherein the neighborhood structure designed in S33 is: S331, adding or deleting a service facility, namely randomly selecting a service facility from a closed service facility set to be started, and randomly selecting a service facility from an open service facility set to be closed; S332, randomly selecting a plurality of pairs of service facilities, wherein each pair of service facilities comprises an open service facility and a closed service facility, exchanging the open and closed states of each pair of service facilities, and after exchanging the states, reallocating the demands according to the mode of S331; s333, reducing the total cost through service facility exchange, namely selecting one closed or open service facility from a closed or open service facility set, and calculating the total cost and unit demand cost before and after the exchange state; S334, randomly opening or closing a plurality of service facilities, wherein under the constraint of budget, waiting time and utilization rate, the probability of selecting a plurality of open service facilities to be closed is 50%, and the probability of selecting a plurality of closed service facilities to be opened is 50%; S335, randomly opening or closing a service facility, wherein the probability of closing the service facility is 50% and the probability of opening the service facility is 50% under the condition that the budget is met; S336, demand distribution, namely screening out open service facilities which are closest to each other in coverage range and have maximum service capacity for unallocated demand points, checking waiting time and utilization rate constraint of the service facilities after new demand is distributed, attempting to promote facility capacity if the constraint is not met, discarding old demand with the demand rate being the lower threshold limit if the constraint is not met, and restoring the service facilities to an initial state if the constraint is not met, budget is exceeded or the service facilities have no demand finally; s337, demand replacement, namely screening open service facilities closest to a coverage area for unallocated demand points, executing demand exchange if the total demand of the service facilities is not greater than the unallocated demand and the demand exchange meets the service quality and utilization rate constraint, otherwise, increasing the number of service desks or improving the technical level; S338, reallocating the unallocated demand, namely screening open service facilities closest to the unallocated demand point in the coverage area, and if the difference value between the service rate and the arrival rate of the service facilities is not smaller than the value corresponding to the demand point and the allocated service facilities meet the constraint of service quality, allocating the demand to the service facilities.
  6. 6. The method for optimizing the location of an electric vehicle battery replacement station taking into account multi-facility synchronization service according to claim 5, wherein the new neighborhood structure generated in S34 is: s341, adding or deleting a service facility, namely randomly selecting a service facility from the closed service facility set and opening the service facility, and randomly selecting a service facility from the open service facility set and closing the service facility; s342, increasing the number of servers, namely randomly selecting an open service facility to increase the number of servers under the condition that budget constraints are met; S343, improving the technical level of the service facilities, namely randomly selecting one service facility to improve the technical level under the condition that budget constraint is met; s344, reassigning the assigned demands, namely randomly selecting one of the assigned demands, reassigning the demand to any one service facility under the constraint of meeting the demand nearby assignment, and checking the waiting time and the utilization rate constraint of the service facility; S345, randomly opening service facilities, namely randomly selecting one service facility from the closed service facility set to be opened, and reallocating allocated and unallocated demands for the opened service facilities; s346, removing specific requirements, namely removing the requirements with the requirement rate being the upper threshold limit for the open service facilities with the utilization rate not less than 0.9, randomly selecting one service facility to reallocate the requirements from other open service facilities in the coverage range of the removed requirements, checking the waiting time and the utilization rate constraint of the new service facilities, and establishing an allocation relation if the constraint is met.

Description

Electric vehicle battery replacement station site selection optimization method considering multi-facility synchronous service Technical Field The invention relates to the technical field of traffic operation management, in particular to an electric automobile power exchange station site selection optimization method considering multi-facility synchronous service. Background At present, the new energy automobile has two energy supplementing modes, namely charging and electricity changing, and the average time of the current electricity changing is 3 minutes or even shorter along with the continuous maturity of the electricity changing technology. The current power conversion mode can shorten the energy supplementing time and improve the vehicle efficiency of a driver. At the same time, the battery replacement demands of the power exchange stations are showing a rapidly increasing trend. It is reported that taxis in many places such as Beijing, sichuan, jiangsu and the like face the phenomenon that the waiting time for power exchange is long, the waiting time for queuing of part of power exchange stations is even more than 1 hour, the power exchange stations are unevenly distributed, the problem is essentially that the explosive growth of demands and the mismatching result of the power exchange stations can reduce the enthusiasm of people on the power exchange technology. At present, most of the power exchange stations are provided with single service stations, but the operation mode is easy to have the phenomena of longer waiting time, less service requirements and the like. Considering only the newly built single service desk power exchange station, the current rapid increase of the power exchange requirement cannot be adapted. There have been studies demonstrating that multiple consoles can serve more needs. Therefore, it is critical to break through the bottleneck to consider the way to use new stations to provide multiple service stations to extend the overall capacity of the urban electricity service network. The improvement of the service capacity and the reasonable layout of the multi-service-desk power exchange station are key to breaking through the bottleneck. Two ways of improving the service capacity of the power exchange station are provided, namely, the battery exchange time is shortened by improving the power exchange technology, but the battery exchange time is difficult to realize in a short period, and finally, the urban power exchange service network is expanded by newly creating the power exchange station, wherein the power exchange station is provided with a plurality of service stations. The construction of the power exchange station has the specificity and complexity, such as the need of a matched high-voltage power transmission network around the construction position of the power exchange station, and the like, and a plurality of alternative positions meeting the construction conditions are generally required to be defined under the discussion of the authorities and expert students. How to select the most suitable construction position from a plurality of alternative positions becomes the first problem to be solved when newly constructing a power exchange station. The existing research is generally developed based on a coverage site selection theoretical model (Covering Location Problem), the M/G/1 and M/G/K queuing theory is applied to site selection, namely, under the constraint that a built power exchange station can cover all the battery replacement requirements of a driver, a site selection model for minimizing the construction cost of the power exchange station is constructed, or the construction budget of the power exchange station is used as a model constraint, a site selection model for maximizing the coverage amount of the battery replacement requirements is constructed, the waiting time of the driver is considered, and finally, a certain algorithm (such as an accurate algorithm, a genetic algorithm, a simulated annealing algorithm and a variable neighborhood search algorithm) is designed to solve the problem model. Although existing research can solve the problem of site selection of a power exchange station in the context of a certain specific problem, a plurality of defects still exist. Disclosure of Invention The invention aims to provide an electric automobile power exchange station site selection optimization method considering multi-facility synchronous service, which solves the problems of long waiting time, uneven distribution and the like of single-facility service in the prior art. In order to achieve the above purpose, the invention provides an electric vehicle battery replacement station site selection optimization method considering multi-facility synchronous service, which comprises the following steps: Step 1, constructing a power exchange station site selection model; step 2, constructing a queuing system and a multi-layer perceptron neural network structu