A method for generating electric vehicle delivery routes considering random demand and time windows

By introducing multiple operators and adaptive mechanisms into the generation method of electric vehicle distribution routes, the distribution route of electric vehicles is optimized, and the problem of electric vehicle paths under the constraints of random demand and time windows is solved, and more accurate energy consumption calculation and cost are achieved. reduce.

CN114492999BActive Publication Date: 2024-11-26武汉禾青优化科技有限公司
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202210095195.7
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2022-01-26
Publication Date
2024-11-26
Estimated Expiration
2042-01-26

AI Technical Summary

Technical Problem

When formulating electric vehicle distribution routes, the prior art ignores the impact of random demand and vehicle load on energy consumption, resulting in the infeasible distribution plan, and the charging demand of electric vehicles is prone to violate customer time windows, increasing distribution costs.

Method used

A method for generating electric vehicle distribution routes that consider random demands and time windows is proposed. By defining multiple removal and insertion operators and adaptive mechanisms, the convergence speed of the solution is optimized, and the number of vehicles and distribution costs are reduced.

Benefits of technology

By accurately calculating the energy consumption of electric vehicles, reducing the number of vehicles, reducing distribution costs, and generating feasible and costly distribution solutions.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN114492999B_ABST
    Figure CN114492999B_ABST
Patent Text Reader

Abstract

The present invention discloses a method for generating an electric vehicle distribution route based on an adaptive large-scale neighborhood search algorithm taking into account random demands and time windows, the method comprising: giving initial data; defining removal and insertion operators; constructing an initial solution; iteratively searching the solution through an adaptive large-scale neighborhood search algorithm; and outputting the optimal solution. The method of the present invention takes into account the impact of random demands on vehicle loads, and can better calculate the energy consumption of electric vehicles and reduce the occurrence of insufficient power. At the same time, an adaptive mechanism is used to adjust the score of the operator, which can effectively improve the convergence speed of the solution. Using this specific implementation method, the generated electric vehicle distribution plan can effectively reduce the number of vehicles required for distribution, thereby reducing distribution costs.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of vehicle routing problem research, and in particular to a method for generating an electric vehicle distribution route taking into account random demand and time windows. Background Art

[0002] The growing demand for transportation has brought tremendous pressure to the environment. Electric vehicles are widely used in distribution activities because they can provide zero tail gas emissions, high efficiency and low noise transportation services. However, there are some limitations in the operation of electric vehicles, such as limited driving range, insufficient number of public charging stations, and long charging time. The vehicle routing problem with time windows (VRPTW) is the core optimization model in the delivery task. Electric vehicles need to be charged in time to overcome their limited driving range. The charging demand of electric vehicles can easily cause customer time window violations, so the electric vehicle routing problem with time windows (EVRPTW) is a very challenging combinatorial optimization problem. At the same time, in order to avoid the situation of insufficient power of electric vehicles during the delivery process, it is necessary to accurately estimate the remaining power of electric vehicles. In existing methods, it is usually assumed that the energy consumption of electric vehicles is proportional to the driving mileage and the impact of vehicle load on energy consumption is ignored.

[0003] However, in the actual distribution process, the uncertainty of customer demand and the limitation of electric vehicle capacity usually lead to the uncertainty of electric vehicle load. The vehicle load is a key factor affecting the energy consumption of electric vehicles. Therefore, it is usually not feasible to ignore the uncertainty of customer demand and the impact of vehicle load on the energy consumption of electric vehicles in the actual distribution process. Therefore, it is crucial to develop a feasible and low-cost distribution plan under the constraints of random demand and time windows, which is also the novelty and creativity of the present invention. Based on this, the present invention proposes a method for generating electric vehicle distribution routes that considers random demand and time windows, defines a variety of removal and insertion operators to improve the solution, and adopts an adaptive mechanism to adjust the score of the operator, which can effectively improve the convergence speed of the solution. Compared with the existing methods, the method proposed in this patent can more accurately calculate the energy consumption of electric vehicles, while reducing the number of vehicles and reducing distribution costs. Summary of the invention

[0004] Aiming at the great complexity of the electric vehicle routing problem with time windows, the present invention provides an electric vehicle distribution route generation method taking into account random demand and time windows to generate an electric vehicle distribution plan and reduce the distribution cost of the enterprise.

[0005] The technical solution of the present invention to solve the above technical problems is as follows:

[0006] An embodiment of the present invention provides a method for generating an electric vehicle delivery route considering random demand and time window, comprising the following steps:

[0007] Step 1: Given customer information, charging station information, the distance matrix between parking lots, customers, and charging stations, and the customer demand distribution and expected delivery time, define the solution S as a set of delivery routes, where each delivery route is a set of ordered customer nodes and charging pile nodes, and define the objective function as minimizing the number of vehicles used and the total power consumption of vehicles. Let n = 0, N NI =25000,N SR =60,N RR =2000;

[0008] Step 2: Define the removal operation including customer node removal operation, charging pile node removal operation, and route removal operation, wherein the customer node removal operation includes the random removal operator: randomly remove K customer nodes in the solution; the minimum target removal operator: remove K customer nodes in the solution to minimize the target function value; the highest similarity removal operator: remove K customers with the highest similarity in the solution; the charging pile node removal operation includes the random removal operator: randomly remove σ charging piles in the solution; the shortest path removal operator: remove σ charging piles in the solution to make the delivery vehicle path shortest; the most charged removal operator: remove the σ charging piles with the most charging times in the solution; the route removal operation includes the random removal operator: randomly remove ω routes in the solution; the minimum number removal operator: remove ω routes with a small number of customers in the solution; the maximum mileage removal operator: remove the ω routes with the longest mileage in the solution; the maximum power consumption removal operator: remove the ω routes with the most power consumption in the solution;

[0009] Step 3: Define the insertion operation including the customer node insertion operation set and the charging pile node insertion operation set, where the customer node insertion operation includes the random insertion operator: randomly inserting customers into the solution; the minimum increment insertion operator: sequentially inserting customers to the position with the minimum increment of the objective function; the minimum energy consumption insertion operator: sequentially inserting customers to the position with the minimum increase in energy consumption; the charging pile node insertion operation includes the minimum distance insertion operator: first find the customer whose battery power is negative when the vehicle arrives, and then insert the charging station in front of the customer so that the distance increase is minimized;

[0010] Step 4: Construct the initial solution S init And assign it to the current solution S cur , the optimal solution S BR , new solution S new ;

[0011] Step 5: If n>N NI Go to step 14, otherwise continue;

[0012] Step 6: If n is divided by N SRIf the remainder is 0, then according to the scores of each operator in the charging pile node removal operation and the charging pile node insertion operation, a charging pile node removal operator and a charging pile node insertion operator are randomly selected through the roulette algorithm, and S is updated according to the selected operators. cur Get temporary solution S temp , go to step 9, otherwise continue;

[0013] Step 7: If n is divided by N RR If the remainder is 0, then according to the scores of each operator in the route removal operation and the customer node insertion operation, a route removal operator and a customer node insertion operator are randomly selected through the roulette algorithm, and S is updated according to the selected operators. cur Get temporary solution S temp , go to step 9, otherwise continue;

[0014] Step 8: According to the scores of each operation in the customer removal operation and the customer node insertion operation, a customer removal operator and a customer node insertion operator are randomly selected through the roulette algorithm, and S is updated according to the selected operators. cur Get temporary solution S temp ;

[0015] Step 9: If S temp If it is not feasible, then for S temp Execute charging pile node insertion operation;

[0016] Step 10: If S temp The objective function is greater than S BR The objective function of S temp After performing a local search, assign the value to S new Otherwise, directly change S temp Assign to S new ;

[0017] Step 11: If S new The objective function is less than S BR The objective function is S new Assign to S BR ;

[0018] Step 12: If S new The objective function is less than S cur The objective function is S new Assign to S cur , otherwise with a certain probability, S new Assign to S cur ;

[0019] Step 13: Update the operator score, set n=n+1, and go to step 5;

[0020] Step 14: Output the optimal solution SBR .

[0021] The embodiment of the present invention provides a method for generating an electric vehicle delivery route taking into account random demand and time windows, which performs iterative search through an adaptive large-scale neighborhood search algorithm; an adaptive mechanism is used to adjust the score of the operator, which can effectively improve the convergence speed of the solution; using the method of this specific implementation, the generated electric vehicle delivery plan can effectively reduce the number of vehicles required for delivery, thereby reducing the delivery cost. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 A schematic diagram of a flow chart of an electric vehicle delivery route generation method considering random demand and time window according to an embodiment of the present invention when being executed;

[0023] Figure 2 Schematic diagram of the initial solution construction process of an embodiment of the present invention. DETAILED DESCRIPTION

[0024] The principles and features of the present invention are described below in conjunction with the accompanying drawings. The examples given are only used to explain the present invention and are not used to limit the scope of the present invention.

[0025] Figure 1 FIG. 1 is a flow chart of an embodiment of the present invention showing a method for generating an electric vehicle delivery route taking into account random demand and time windows when being executed. Figure 1 As shown, the following steps are included:

[0026] Step 1: Given customer information, charging station information, the distance matrix between parking lots, customers, and charging stations, and the customer demand distribution and expected delivery time, define the solution S as a set of delivery routes, where each delivery route is a set of ordered customer nodes and charging pile nodes, and define the objective function as minimizing the number of vehicles used and the total power consumption of vehicles. Let n = 0, N NI =25000,N SR =60,N RR =2000;

[0027] Step 2: Define the removal operation including customer node removal operation, charging pile node removal operation, and route removal operation, wherein the customer node removal operation includes the random removal operator: randomly remove K customer nodes in the solution; the minimum target removal operator: remove K customer nodes in the solution to minimize the target function value; the highest similarity removal operator: remove K customers with the highest similarity in the solution; the charging pile node removal operation includes the random removal operator: randomly remove σ charging piles in the solution; the shortest path removal operator: remove σ charging piles in the solution to make the delivery vehicle path shortest; the most charged removal operator: remove the σ charging piles with the most charging times in the solution; the route removal operation includes the random removal operator: randomly remove ω routes in the solution; the minimum number removal operator: remove ω routes with a small number of customers in the solution; the maximum mileage removal operator: remove the ω routes with the longest mileage in the solution; the maximum power consumption removal operator: remove the ω routes with the most power consumption in the solution;

[0028] Step 3: Define the insertion operation including the customer node insertion operation set and the charging pile node insertion operation set, where the customer node insertion operation includes the random insertion operator: randomly inserting customers into the solution; the minimum increment insertion operator: sequentially inserting customers to the position with the minimum increment of the objective function; the minimum energy consumption insertion operator: sequentially inserting customers to the position with the minimum increase in energy consumption; the charging pile node insertion operation includes the minimum distance insertion operator: first find the customer whose battery power is negative when the vehicle arrives, and then insert the charging station in front of the customer so that the distance increase is minimized;

[0029] Step 4: Construct the initial solution S init And assign it to the current solution S cur , the optimal solution S BR , new solution S new ;

[0030] Step 5: If n>N NI Go to step 14, otherwise continue;

[0031] Step 6: If n is divided by N SR If the remainder is 0, then according to the scores of each operator in the charging pile node removal operation and the charging pile node insertion operation, a charging pile node removal operator and a charging pile node insertion operator are randomly selected through the roulette algorithm, and S is updated according to the selected operators. cur Get temporary solution S temp , go to step 9, otherwise continue;

[0032] Step 7: If n is divided by N RR If the remainder is 0, then according to the scores of each operator in the route removal operation and the customer node insertion operation, a route removal operator and a customer node insertion operator are randomly selected through the roulette algorithm, and S is updated according to the selected operators. cur Get temporary solution Stemp , go to step 9, otherwise continue;

[0033] Step 8: According to the scores of each operation in the customer removal operation and the customer node insertion operation, a customer removal operator and a customer node insertion operator are randomly selected through the roulette algorithm, and S is updated according to the selected operators. cur Get temporary solution S temp ;

[0034] Step 9: If S temp If it is not feasible, then for S temp Execute charging pile node insertion operation;

[0035] Step 10: If S temp The objective function is greater than S BR The objective function of S temp After performing a local search, assign the value to S new Otherwise, directly change S temp Assign to S new ;

[0036] Step 11: If S new The objective function is less than S BR The objective function is S new Assign to S BR ;

[0037] Step 12: If S new The objective function is less than S cur The objective function is S new Assign to S cur , otherwise with a certain probability, S new Assign to S cur ;

[0038] Step 13: Update the operator score, set n=n+1, and go to step 5;

[0039] Step 14: Output the optimal solution S BR .

[0040] Specifically, the mathematical expression of the objective function in step 1 is , that is, minimize the number of vehicles used and minimize the electricity consumption cost of all vehicles, where x ij is a binary decision variable, which takes 1 when the vehicle visits arc (i, j) and 0 otherwise. 0j Represents the exit arc of the vehicle, node 0 represents the parking lot, and C ij (q) represents the power consumption of the vehicle when it passes through arc (i, j)∈A when the customer demand is q, C F With C ERepresent the unit vehicle cost and unit electricity cost respectively. The criteria for judging whether the solution is feasible in step 9 include: (1) the total transportation demand of each route in the solution cannot exceed the vehicle capacity; (2) it can meet the time window requirements of all customer nodes, that is, the time to arrive at the customer node is within the acceptable range for passengers; (3) it does not violate the upper limit of the vehicle's power, that is, the electric vehicle's power will not fall below the minimum value. In step 9, S temp After inserting the customer node, it is possible to violate the power limit, so it is necessary to perform the charging pile node insertion operation to make it feasible.

[0041] Figure 2 Schematic diagram of the initial decomposition process of an embodiment of the present invention, as shown in Figure 2 As shown, the following steps are included:

[0042] Step 41: Initialize the unassigned customer set C R is the set of all client nodes, the initial solution S init is empty;

[0043] Step 42: Select the customer farthest from the parking lot and move him from C R Remove it and assign it a new line k;

[0044] Step 43: Determine set C R Is it empty? If it is empty, go to step 47;

[0045] Step 44: In Set C R Select a customer node c that minimizes the cost increase of the current line i ;

[0046] Step 45: Determine c i After inserting k, whether the time window and capacity constraints are met, if yes, c i Insert k and move it from C R Otherwise, insert line k into S init , and c i Allocate a new line k=k+1, and go to step 43;

[0047] Step 46: Determine whether the current line needs to be charged. If not, go to step 43; otherwise, insert a charging pile node in k and go to step 43;

[0048] Step 47: Output the initial solution S init .

[0049] Specifically, the time window and capacity constraints in step 45 refer to: (1) client node c i After inserting into route k, the vehicle arrives at c i The time cannot be earlier than customer ci The earliest delivery time required cannot be later than the customer c i Required latest delivery time; (2) Customer node c i After inserting route k, the maximum demand of all customers on the route cannot exceed the capacity of the electric vehicle. This method is a heuristic method that can provide a feasible initial solution for the algorithm.

[0050] Simulation results and analysis

[0051] In order to verify the effectiveness of this specific implementation method, this section conducts experiments on the E-VRPTW problem with determined demand and considering the energy consumption model. The experimental example selects a test case with a scale of 100. Each test case is run 10 times with the same algorithm parameters, and the solution with a small number of corresponding vehicles is preferred. The comparison results of the method proposed in this patent and the existing methods are shown in Table 1. The m in the table IC Represents the number of traditional fuel vehicles, m E represents the number of electric vehicles, f d Represents the total mileage in kilometers.

[0052] Table 1 A brief comparison of this specific implementation method with existing methods for a case with a scale of 100

[0053]

[0054] As can be seen from Table 1, compared with the solution given by the traditional method, the solution of this specific implementation method uses a smaller number of vehicles, but due to the need to charge electric vehicles, the total mileage of the solution is greater.

[0055] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, i.e., they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Those of ordinary skill in the art may understand and implement it without creative effort.

[0056] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0057] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for generating an electric vehicle delivery route considering random demand and time window, characterized in that: A reasonable set of electric vehicle delivery routes can be generated according to customer needs and delivery time requirements. The specific steps are as follows: Step 1: Given customer information, charging station information, the distance matrix between parking lots, customers, and charging stations, and the customer demand distribution and expected delivery time, define the solution S as a set of delivery routes, where each delivery route is a set of customer nodes and charging pile nodes with order, and define the objective function as minimizing the number of vehicles used and the total power consumption of vehicles. Let n = 0, N NI =25000,N SR =60,N RR =2000; Step 2: Define the removal operation including customer node removal operation, charging pile node removal operation, and route removal operation. The customer node removal operation includes the random removal operator: randomly remove K customer nodes in the solution; the minimum target removal operator: remove K customer nodes in the solution to minimize the target function value; the highest similarity removal operator: remove the K customers with the highest similarity in the solution; the charging pile node removal operation includes the random removal operator: randomly remove σ charging piles in the solution; the shortest path removal operator: remove σ charging piles in the solution to make the delivery vehicle path shortest; the most charged removal operator: remove the σ charging piles with the most charging times in the solution; The route removal operation includes the random removal operator: randomly removing ω routes from the solution; Minimum number removal operator: removes ω routes with a small number of customers in the solution; Maximum mileage removal operator: remove the ω routes with the longest mileage in the solution; The operator with the highest power consumption is removed: the ω routes with the highest power consumption in the solution are removed; Step 3: Define the insertion operation including a customer node insertion operation set and a charging pile node insertion operation set, wherein the customer node insertion operation includes a random insertion operator: randomly inserting the customer into the solution; Minimum increment insertion operator: insert customers to the position with the smallest increment of the objective function in sequence; Minimum energy consumption insertion operator: insert customers to the position with the smallest increase in energy consumption in sequence; Charging pile node insertion operation includes minimum distance insertion operator: first find the customer whose battery power is negative when the vehicle arrives, and then insert the charging station in front of the customer to minimize the increase in distance; Step 4: Construct the initial solution S init And assign it to the current solution S cur , the optimal solution S BR , new solution S new ; Specifically: Step 41: Initialize the unassigned customer set C R is the set of all client nodes, the initial solution S init is empty; Step 42: Select a customer who is farthest from the parking lot and remove him from C R Remove it and assign it a new line k; Step 43: Determine set C R Is it empty? If it is empty, go to step 47; Step 44: In set C R Select a customer node ci that minimizes the cost increase of the current line; Step 45: Determine c i After inserting k, whether the time window and capacity constraints are met, if so, insert ci into k and remove it from C R Otherwise, insert line k into S init , and c i Assign a new line k=k+1, and go to step 43; Step 46: Determine whether the current line needs to be charged. If not, go to step 43; otherwise, insert a charging pile node in k, and go to step 43; Step 47: Output the initial solution S init , let the current solution S cur , the optimal solution S BR , new solution S new , which is equal to the initial solution S init ; Step 5: If n>N NI Go to step 14, otherwise continue; Step 6: If n is divided by N SR If the remainder is 0, then according to the scores of each operator in the charging pile node removal operation and the charging pile node insertion operation, a charging pile node removal operator and a charging pile node insertion operator are randomly selected through the roulette algorithm, and S is updated according to the selected operators. cur Get temporary solution S temp , go to step 9, otherwise continue; Step 7: If n is divided by N RR If the remainder is 0, then according to the scores of each operator in the route removal operation and the customer node insertion operation, a route removal operator and a customer node insertion operator are randomly selected through the roulette algorithm, and S is updated according to the selected operators. cur Get temporary solution S temp , go to step 9, otherwise continue; Step 8: According to the scores of each operation in the customer removal operation and the customer node insertion operation, a customer removal operator and a customer node insertion operator are randomly selected through the roulette algorithm, and S is updated according to the selected operators. cur Get temporary solution S temp ; Step 9: If S temp If this is not feasible, then for S temp Execute the charging pile node insertion operator; Step 10: If S temp The objective function is greater than S BR The objective function of S temp After performing a local search, assign the value to S new Otherwise, directly change S temp Assign to S new ; Step 11: If S new The objective function is less than S BR The objective function is S new Assign to S BR ; Step 12: If S new The objective function is less than S cur The objective function is S new Assign to S cur , otherwise with a certain probability, S new Assign to S cur ; Step 13: Update the operator score, set n = n + 1, and go to step 5; Specifically: Step 131: If the new solution S new The objective function is less than the optimal solution S BR The objective function is, then the score of the selected operator is increased by 25; Step 132: If the new solution S new The objective function is less than the current solution S cur The objective function of the selected operator is increased by 21; Step 133: If the new solution S new The objective function is greater than the optimal solution S BR With the current solution S cur The objective function of , but it is still selected by the algorithm and assigned to S cur , then the selected operator score increases by 20; Step 14: Output the optimal solution S BR .

2. The method for generating an electric vehicle delivery route considering random demand and time window according to claim 1, characterized in that: The criteria for judging whether a solution is feasible include: (1) The maximum transport demand of each route in the solution cannot exceed the vehicle capacity; (2) It can meet the time window requirements of all customer nodes, that is, the time to arrive at the customer node is within the acceptable range for passengers; (3) During transportation, the battery level of the vehicle must not be less than the minimum level.