Multi-center hybrid fleet joint scheduling method based on improved genetic variable neighborhood algorithm

Abstract

The invention provides a multi-center hybrid fleet joint scheduling method based on an improved genetic variable neighborhood algorithm, and the method comprises the steps: building a joint scheduling model with the minimization of vehicle transportation cost, charging cost and overtime penalty cost as a target, and building constraint conditions; a genetic algorithm is adopted for optimization, and a joint scheduling scheme with the lowest total cost under the constraint condition is obtained; performing variable neighborhood local search on each generation of optimal individuals after genetic manipulation; executing a dynamic disturbance operation; and judging whether the maximum number of iterations is reached or not, if so, ending the process, and obtaining a multi-center hybrid fleet joint scheduling scheme, otherwise, repeating the previous steps. According to the invention, a joint scheduling mechanism is introduced to realize transportation resource sharing, and the overall resource configuration is optimized; a joint scheduling model is constructed based on a hybrid motorcade, an electric vehicle charging strategy is optimally designed, an improved genetic algorithm and a variable neighborhood algorithm are used for optimization solution, and customer service satisfaction experience can be improved.

Description

Technical Field

[0001] The present invention relates to the technical field of vehicle path optimization, and in particular to a multi-center mixed vehicle fleet joint scheduling method based on an improved genetic variable neighborhood algorithm. Background Art

[0002] In urban logistics distribution scenarios operated by multiple distribution centers, a mixed fleet of electric and fuel-powered vehicles is used for delivery operations. However, due to limited battery capacity, electric vehicles face insufficient range during delivery. Recharging is necessary to extend their range, but prolonged charging times can lead to delays in delivery and a poor customer service experience. Therefore, optimizing electric vehicle charging strategies is being considered, allowing for partial charging at charging stations. This ensures sufficient charge to complete the remaining delivery, while minimizing charging time and minimizing the impact on delivery journeys. Furthermore, with increasing customer demand, the conflict between limited transportation resources and growing customer demands is becoming increasingly prominent. Therefore, joint scheduling is being implemented across multiple distribution centers, allowing vehicle transportation resources to be shared across centers and reducing the costs and resource waste associated with idle return trips. These measures optimize overall resource allocation, improve delivery service efficiency, and reduce operating costs.

[0003] In the logistics network, there are three types of logistics nodes, namely logistics distribution center D, customer point C and charging station R. N = D∪C∪R represents the set of all points, and the arc set L = {(i,j)|i,j∈N}. The vehicle set of each logistics distribution center is represented by V d =V ed ∪V od Denotes, where d∈D, V ed and V od Represents a collection of electric and fuel-powered vehicles. During the delivery process, the customer's freight volume, distance, delivery window, and duration are known. Each logistics distribution center dispatches vehicles to meet the customer's freight needs. All vehicles depart from each logistics distribution center and, upon completion of a delivery, do not need to return to their original distribution center; instead, they can dock at any nearby distribution center. Electric vehicles have a minimum charge threshold; if their charge runs low during a journey, they must recharge at a charging station. Furthermore, vehicles are not permitted to exceed their service time, as this incurs a penalty fee.

[0004] In the related art, a hybrid taboo search algorithm is used to combine the taboo search algorithm, variable neighborhood search, artificial immune algorithm and large neighborhood search, etc., so as to effectively utilize the advantages of various algorithms for local search and global search, thereby reducing the number of vehicles used and reducing costs. However, it does not fully consider the joint scheduling of different types of vehicles and based on different logistics nodes in this application. Summary of the Invention

[0005] The purpose of the present invention is to provide a logistics joint scheduling method for different types of vehicles and multiple types of logistics nodes to achieve transportation resource sharing and optimize overall resource allocation in response to the shortcomings of the above-mentioned background technology.

[0006] In order to achieve the above objectives, the present invention provides a multi-center mixed fleet joint scheduling method based on an improved genetic variable neighborhood algorithm. The logistics network involved contains three types of logistics nodes consisting of logistics distribution centers D, customer points C, and charging stations R. N = D∪C∪R represents the set of all points, the arc set L = {(i, j)|i, j∈N}, and the vehicle set of the logistics distribution center is represented by V d =V ed ∪V od Denotes, d∈D, V ed and V od Represents the collection of electric vehicles and fuel vehicles;

[0007] The steps include:

[0008] S1, establish a joint scheduling model with the goal of minimizing vehicle transportation cost, charging cost and overtime penalty cost, and establish constraints at the same time;

[0009] S2, uses genetic algorithm to optimize and obtain the joint scheduling solution with the lowest total cost under the constraints of S1;

[0010] S3, perform variable neighborhood local search for the best individual of each generation after genetic operation;

[0011] S4, perform dynamic perturbation operation;

[0012] S5, determine whether the maximum number of iterations has been reached. If so, the process ends and a multi-center mixed fleet joint scheduling solution is obtained. Otherwise, repeat S3 to S4.

[0013] Furthermore, in the joint scheduling model, in the distribution link, the freight volume, distance, delivery time window, and delivery duration of the customer point are known, and each logistics distribution center dispatches vehicles to meet the customer's freight needs; all vehicles depart from each logistics distribution center and do not need to return to the original logistics distribution center after the freight task is completed, but can stop at any nearby logistics distribution center; electric vehicles have a minimum power threshold, and when the power is insufficient during the journey, they need to go to the charging station to replenish electricity; vehicles are not allowed to serve customers overtime, otherwise penalty fees will be incurred.

[0014] Furthermore, in S1, the minimum total cost F consisting of vehicle transportation cost F1, charging cost F2 and overtime penalty cost F3 is used as the objective function, and the joint scheduling model is established as follows:

[0015] minF=F1+F2+F3

[0016] Among them, the vehicle transportation cost F1 is expressed by the following formula:

[0017]

[0018] The charging cost F2 is expressed by the following formula:

[0019]

[0020] The overtime penalty cost F3 is expressed by the following formula:

[0021]

[0022] Among them, f1 and f2 represent the unit transportation costs of electric vehicles and fuel vehicles respectively; d ij represents the distance of arc (i, j); represents a 0-1 variable, which is 1 if the vehicle v passes through the arc (i, j) and 0 if not; represents the charging amount of electric vehicle v at charging station r; f3, f4 represent the unit charging cost and unit time penalty cost respectively; b i represents the time window of customer i; α iv Indicates the time when the vehicle arrives.

[0023] Furthermore, the constraints set in S1 include: each customer is delivered by the vehicle only once; the load of electric vehicles and fuel vehicles is not greater than their respective capacities; the vehicles comply with the cargo flow balance constraint; the vehicles can stop at any logistics distribution center on the return trip; the vehicle flow balance constraint at the customer point; the electric vehicle is charged at most once; the number of vehicles in each center is limited; the relationship between the power of the electric vehicle before and after passing the arc (i, j); the linear relationship between the charging amount and the charging rate; the minimum power limit constraint of the electric vehicle; the power of the electric vehicle after charging must be able to meet the remaining travel requirements; the electric vehicle starts with a full charge; the vehicle leaves from point i, and if it passes through arc (i, j), the time to arrive at point j is greater than or equal to the departure time of point i plus the travel time on arc (i, j); the time constraint for the vehicle to arrive at and leave the customer point and the charging station.

[0024] Furthermore, S2 specifically includes the following sub-steps:

[0025] S21, set the initial parameters, including the maximum number of iterations, population size, crossover rate, and mutation rate;

[0026] S22, uses greedy clustering to construct the population to generate a high-quality initial population;

[0027] S23, taking the inverse of the objective function F (1 / F) as the fitness function and performing fitness calculation;

[0028] S24, sort the individuals in the population from high to low according to their fitness, and add the preset proportion of individuals with the highest ranking to the offspring population;

[0029] S25, random order crossover was performed;

[0030] S26, performs parallel mutation;

[0031] S27, for EV routes with infeasible power, a backtracking method is used to insert charging stations and repair infeasible solutions to meet actual constraints.

[0032] Furthermore, S22 includes the following sub-steps:

[0033] S221 divides all customers into corresponding logistics distribution centers according to the principle of closest distance, and obtains multiple unit clusters after clustering each logistics distribution center;

[0034] In the S222 unit cluster, a customer is randomly selected and an initial solution path that satisfies the vehicle capacity constraint is constructed according to the greedy criterion;

[0035] S223, repeat S222 until the population individuals reach the scale.

[0036] Furthermore, S27 includes the following sub-steps:

[0037] S271, checking the circuit to determine the position of the breakpoint where the power is insufficient;

[0038] S272, inserting charging stations with the shortest distances to the front and rear customers in the interval from the breakpoint position in sequence from the back to the front;

[0039] S273, calculating the incremental cost of the line after each charging station is inserted, and inserting the charging station at the position with the minimum incremental cost to obtain a feasible solution after repair.

[0040] Furthermore, the neighborhood structure N is designed in S3 k , which includes 6 different operation operators, namely N1Swap: swap the positions of two customers in the current path; N22-opt: reverse the middle segment of two customers in the current path; N3InterSwap1-1: select a customer on each of the two paths for exchange; N41-2-3insert: select 1, 2, and 3 customers from the three paths respectively and remove them, and then reinsert them into the corresponding paths in the order of 1→2, 2→3, and 3→1; N5DepotSwap1-1: select a customer on each path in the two distribution centers for exchange; N6DeleteSmallestRoute: delete the path with the least customers and put them back into the remaining paths.

[0041] Furthermore, S3 includes the following sub-steps:

[0042] S31, input the current solution x, set k = 1;

[0043] S32, using the neighborhood structure N k When performing local search, the main thread distributes the current solution x to the first k neighborhood structures at the same time, and each neighborhood structure acts as a thread to generate candidate solutions x in the current neighborhood. k And calculate the corresponding target value;

[0044] S33, the main thread compares the target values ​​of all candidate solutions and selects the best candidate solution x best ;

[0045] S34, the main thread compares the best candidate solution x best and the current solution x target value, and determine whether an improvement is obtained. If so, update the current solution to the optimal candidate solution and return to S31 to continue searching; otherwise, keep the current solution unchanged and set k = k + 1;

[0046] S35, check whether the neighborhood search termination condition is met. If k≤6, return to S32 and restart the search. Otherwise, end the search process and enter S4.

[0047] Furthermore, S4 includes the following sub-steps:

[0048] S41, is the neighborhood structure N k The six operators included are given the same weight so that each operator has the same probability of being selected;

[0049] S42, during the perturbation process, evaluates the quality improvement effect of the solution before and after the application of each operator;

[0050] S43, dynamically assign new weights to operators based on the improvement effects, including increasing the weights of operators with outstanding improvement effects and reducing the weights of operators with poor performance. Operators with excellent performance are given priority in subsequent disturbances.

[0051] The above solution of the present invention has the following beneficial effects:

[0052] The multi-center mixed fleet joint scheduling method based on the improved genetic variable neighborhood algorithm provided by the present invention takes into account the scenario where multiple logistics distribution centers use mixed fleets to carry out logistics distribution operations, introduces a joint scheduling mechanism to achieve transportation resource sharing, and optimizes overall resource allocation. A joint scheduling model is constructed based on the mixed fleet, and the electric vehicle charging strategy is optimized and designed. At the same time, the improved genetic algorithm and variable neighborhood algorithm are used for optimization and solution, which is conducive to improving customer service satisfaction experience. Compared with existing technologies, the joint scheduling can be more reasonable and efficient.

[0053] Other beneficial effects of the present invention will be described in detail in the subsequent specific implementation section. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 is a flow chart of the steps of the present invention;

[0055] Figure 2 Schematic diagram of the mixed fleet joint dispatching operation of the present invention. DETAILED DESCRIPTION

[0056] The following describes the embodiments of the present disclosure through specific examples, and those skilled in the art can easily understand other advantages and effects of the present disclosure from the contents disclosed in this specification. Obviously, the described embodiments are only a part of the embodiments of the present disclosure, rather than all of the embodiments. The present disclosure can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present disclosure. It should be noted that, in the absence of conflict, the following embodiments and features in the embodiments can be combined with each other. Based on the embodiments in the present disclosure, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present disclosure.

[0057] It should be noted that various aspects of the embodiments within the scope of the appended claims are described below. It should be apparent that the aspects described herein can be embodied in a wide variety of forms, and any specific structure and / or function described herein is merely illustrative. Based on this disclosure, it should be understood by those skilled in the art that an aspect described herein can be implemented independently of any other aspect, and two or more of these aspects can be combined in various ways. For example, any number of aspects described herein can be used to implement an apparatus and / or practice a method. In addition, other structures and / or functionalities other than one or more of the aspects described herein can be used to implement this apparatus and / or practice this method.

[0058] It should also be noted that the diagrams provided in the following embodiments are merely schematic illustrations of the basic concepts of the present disclosure. The diagrams only show components relevant to the present disclosure and are not drawn according to the number, shape, and size of components in actual implementation. In actual implementation, the configuration, quantity, and proportion of each component may be varied at will, and the component layout may be more complex. Furthermore, in the following description, specific details are provided to facilitate a thorough understanding of the examples. However, those skilled in the art will appreciate that the described aspects may be practiced without these specific details.

[0059] like Figure 1As shown, the embodiment of the present invention provides a multi-center mixed fleet joint scheduling method based on an improved genetic variable neighborhood algorithm. The problem involved is that the logistics network contains three types of logistics nodes consisting of logistics distribution centers D, customer points C, and charging stations R. N = D∪C∪R represents the set of all points, and the arc set L = {(i, j)|i, j∈N}. The vehicle set of each logistics distribution center is represented by V d =V ed ∪V od Denotes, where d∈D, V ed and V od Represents a collection of electric and fuel-powered vehicles. During the delivery process, the customer's freight volume, distance, delivery window, and duration are known. Each logistics distribution center dispatches vehicles to meet the customer's freight needs. All vehicles depart from each logistics distribution center and, upon completion of a delivery, do not need to return to their original distribution center; instead, they can dock at any nearby distribution center. Electric vehicles have a minimum charge threshold; if their charge runs low during a journey, they must recharge at a charging station. Furthermore, vehicles are not permitted to exceed their service time, as this incurs a penalty fee.

[0060] Therefore, in S1, a joint scheduling model is established with the goal of minimizing vehicle transportation cost, charging cost and overtime penalty cost, and constraints are set at the same time.

[0061] For example, Figure 2 As shown in the figure, three logistics distribution centers each dispatched a vehicle to provide freight services to 12 customers. Logistics distribution center 1 dispatched a fuel truck, which served customers 10 and 8, then docked at logistics distribution center 3. Logistics distribution center 2 dispatched two fuel trucks and an electric vehicle. Fuel truck 1 served customers 3 and 12, then docked at logistics distribution center 3. Fuel truck 2 served customers 7, 2, and 5, then returned to the original logistics distribution center. The electric vehicle, after serving customers 6 and 9, charged at a charging station due to low battery, then continued to serve customer 11, docking at distribution center 1. Distribution center 3 dispatched an electric vehicle, which served customers 4 and 1, then returned to the original center.

[0062] Therefore, taking the minimum total cost F consisting of vehicle transportation cost F1, charging cost F2 and overtime penalty cost F3 as the objective function, the joint scheduling model is established as follows:

[0063] minF=F1+F2+F3(1)

[0064] Among them, the vehicle transportation cost F1 can be expressed by the following formula:

[0065]

[0066] The charging cost F2 can be expressed by the following formula:

[0067]

[0068] The overtime penalty cost F3 can be expressed by the following formula:

[0069]

[0070] At the same time, establish the following constraints:

[0071]

[0072]

[0073] δ iv ≥B e ·Soc (15)

[0074]

[0075] θ iv =B e (17)

[0076]

[0077] Among them, formula (5) means that each customer is delivered by the vehicle only once; formulas (6) and (7) respectively mean that the load of electric vehicles and fuel vehicles is not greater than their respective capacities; formula (8) means that the vehicle complies with the cargo flow balance constraint; formula (9) means that the vehicle can stop at any logistics distribution center when returning; formula (10) means that the vehicle flow balance at the customer point; formula (11) means that the electric vehicle can be charged at most once; formula (12) means the number of vehicles in each center is limited; formula (13) means the relationship between the electric vehicle's power before and after passing arc (i, j); formula (14) means the charging amount It is linearly related to the charging rate; Formula (15) represents the minimum power limit of the electric vehicle; Formula (16) represents that the power of the electric vehicle after charging must be able to meet the remaining travel requirements; Formula (17) represents that the electric vehicle starts with a full charge; Formula (18) represents that if the vehicle leaves from point i and passes through arc (i, j), the time to arrive at point j is greater than or equal to the departure time of point i plus the travel time on arc (i, j); Formulas (19) and (20) represent the time constraints for the vehicle to arrive at and leave the customer point and the charging station respectively; Formulas (21) to (23) represent the type and value constraints of the variables. The meanings of the parameters in the above formulas are shown in Table 1:

[0078] Table 1 Parameter meaning

[0079]

[0080]

[0081] The subsequent steps of the multi-center mixed fleet joint scheduling method based on the improved genetic variable neighborhood algorithm provided in this embodiment include:

[0082] S2, using genetic algorithm (GA) to optimize, to obtain the joint scheduling solution with the lowest total cost under the constraints of S1. This step specifically includes the following sub-steps:

[0083] S21: Set initial parameters, including the maximum number of iterations, population size, crossover rate, and mutation rate. This step initializes the algorithm hyperparameters.

[0084] S22, greedy clustering is used to construct the population to generate a high-quality initial population.

[0085] Specifically in this embodiment, this step includes the following sub-steps: S221 divides all customers into corresponding logistics distribution centers according to the principle of shortest distance, and obtains multiple unit clusters after clustering each logistics distribution center; S222 randomly selects a customer from the unit cluster, and constructs an initial solution path that meets the vehicle capacity limit according to the greedy criterion; S223, repeats S222 until the population individuals reach the scale.

[0086] S23, taking the inverse of the objective function F (1 / F) as the fitness function, and performing fitness calculation to quantify the individual's quality.

[0087] S24, sort the individuals in the population from high to low according to their fitness, and add the top 20% of individuals to the offspring population to achieve elite strategy selection. This step is the selection operation of the genetic algorithm to retain the elite individuals.

[0088] S25, perform random order crossover. This step is the crossover operation of the legacy algorithm, which recombines genes to generate new individuals.

[0089] Specifically in this embodiment, this step includes the following sub-steps: S251, generating two sets of random numbers with the same length as the chromosome between the interval 0 and 1, corresponding to the point genes on the parents 1 and 2 respectively; S252, comparing the random number corresponding to each point with the crossover rate, and retaining the chromosome point genes greater than the crossover rate; S253, assigning the retained point genes in the parent 1 to the offspring 2, and then copying the genes lacking in the offspring 2 from the parent 2 in sequence; S254, repeating S253 to obtain another offspring individual.

[0090] S26, performing parallel mutation. This step is the mutation operation of the genetic algorithm, which introduces random perturbations to avoid premature maturation.

[0091] Specifically in this embodiment, this step includes the following sub-steps: S261, randomly deleting some point genes on the chromosome and performing permutations and combinations; S262, reinserting the sub-permutations in the combination into the individuals with deleted genes; S263, calculating the fitness values ​​of all new individuals in parallel and comparing them; S264, retaining the individual corresponding to the maximum value.

[0092] S27, for EV routes with infeasible power, a backtracking method is used to insert charging stations and repair infeasible solutions to meet actual constraints.

[0093] Specifically in this embodiment, this step includes the following sub-steps: S271, inspecting the line to determine the location of the breakpoint where the power is insufficient; S272, inserting charging stations with the smallest distance to the front and rear customers in the interval from the breakpoint location from back to front; S273, calculating the incremental cost of the line after each charging station is inserted, and inserting the charging station at the location with the smallest incremental cost to obtain a feasible solution after repair.

[0094] S3, performs variable neighborhood local search on the best individual of each generation after genetic operation.

[0095] In this embodiment, in order to accelerate convergence, a parallel mechanism is introduced in the search process to improve efficiency. Specifically, the neighborhood structure N is designed. k , which includes 6 different operation operators, namely N1Swap: swap the positions of two customers in the current path; N22-opt: reverse the middle segment of two customers in the current path; N3InterSwap1-1: select a customer on each of the two paths for exchange; N41-2-3insert: select 1, 2, and 3 customers from the three paths respectively and remove them, and then reinsert them into the corresponding paths in the order of "1→2, 2→3, 3→1"; N5DepotSwap1-1: select a customer on each path in the two distribution centers for exchange; N6DeleteSmallestRoute: delete the path with the least customers and put them back into the remaining paths.

[0096] Based on these six different operators, the specific operation method of the variable neighborhood search is as follows: S31, input the current solution x, set k = 1; S32, use the neighborhood structure N k When performing local search, the main thread distributes the current solution x to the first k neighborhood structures at the same time, and each neighborhood structure acts as a thread to generate candidate solutions x in the current neighborhood. k And calculate the corresponding target value; S33, the main thread compares the target values ​​of all candidate solutions and selects the best candidate solution x best ; S34, the main thread compares the optimal candidate solution x bestand the current solution x target value to determine whether improvement is achieved. If so, update the current solution to the optimal candidate solution and return to S31 to continue searching. Otherwise, keep the current solution unchanged and set k = k + 1. In S35, check whether the neighborhood search termination condition is met. If k ≤ 6, return to S32 and restart the search. Otherwise, end the search process and enter S4.

[0097] S4, performs a dynamic perturbation operation. Perturbation involves selecting one operation from the six neighborhood structures each time to expand the solution space. During perturbation, a dynamic weighting strategy is introduced to further improve search efficiency and solution quality.

[0098] Specifically in this embodiment, the step includes: S41, firstly, the neighborhood structure N k The six included operation operators are assigned the same weight so that each operator has the same probability of being selected; S42, during the perturbation process, evaluates the quality improvement effect of the solution before and after the application of each operator; S43, dynamically assigns new weights to operators based on the improvement effect, including increasing the weights of operators with outstanding improvement effects and reducing the weights of operators with poor performance, to ensure that operation operators with excellent performance are given priority in subsequent perturbations.

[0099] S5, determine whether the maximum number of iterations has been reached. If so, the process ends and a multi-center mixed fleet joint scheduling solution is obtained. Otherwise, repeat S3 to S4.

[0100] As described above, this method considers the scenario where multiple logistics distribution centers use mixed fleets to carry out logistics distribution operations, introduces a joint scheduling mechanism to realize transportation resource sharing, and optimizes the overall resource allocation; a joint scheduling model is constructed based on the mixed fleet, and the electric vehicle charging strategy is optimized and designed. At the same time, an improved genetic variable neighborhood algorithm is used for optimization and solution, which is conducive to improving customer service satisfaction experience. Compared with existing technologies, joint scheduling can be more reasonable and efficient.

[0101] The effectiveness of this method is further illustrated below through real-world examples. Simulation tests using MATLAB software show that for a case involving three distribution centers, five charging stations, and 60 customer locations, the method provided by this embodiment achieves the optimal solution within 54 seconds, and the average error between the optimal value and the average value across five independent tests is 1.62%. For a case involving five distribution centers, six charging stations, and 80 customer locations, the method provided by this embodiment achieves the optimal solution within 82 seconds, and the average error between the optimal value and the average value across five independent tests is 1.89%. These results demonstrate that this method is effective for solving small-scale multi-center mixed fleet joint scheduling problems.

[0102] Based on the same inventive concept, this embodiment also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the aforementioned multi-center mixed fleet joint scheduling method based on the improved genetic variable neighborhood algorithm.

[0103] The computer-readable medium includes, but is not limited to, any type of disk (including floppy disks, hard disks, optical disks, CD-ROMs, and magneto-optical disks), ROMs, RAMs, EPROMs (Erasable Programmable Read-Only Memory), EEPROMs, flash memories, magnetic cards, or optical cards. In other words, the computer-readable medium includes any medium that can store or transmit information in a form that can be read by a device (e.g., a printer).

[0104] Based on the same inventive concept, this embodiment further provides a device comprising at least one processor and a memory communicatively coupled to the at least one processor. The memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute the aforementioned multi-center mixed fleet joint scheduling method based on the improved genetic variable neighborhood algorithm.

[0105] The computer-readable storage medium, device, etc. provided in this embodiment have the same inventive concept and the same beneficial effects as the aforementioned method, and will not be described in detail here.

[0106] The technical features of the above embodiments can be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0107] The above embodiments merely illustrate several implementation methods of the present application. While the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the present application. It should be noted that a person of ordinary skill in the art may make various modifications and improvements without departing from the spirit of the present application, all of which fall within the scope of protection of the present application. Therefore, the scope of protection of the present application shall be determined by the appended claims.

Claims

1. A multi-center mixed fleet joint scheduling method based on an improved genetic variable neighborhood algorithm. The logistics network involved contains three types of logistics nodes: logistics distribution center D, customer points C, and charging stations R. N = D∪C∪R represents the set of all points, the arc set L = {(i,j)|i,j∈N}, and the vehicle set of the logistics distribution center is represented by V d =V ed ∪V od Denotes, d∈D, V ed and V od Represents the collection of electric vehicles and fuel vehicles; It is characterized by: The steps include: S1, establish a joint scheduling model with the goal of minimizing vehicle transportation cost, charging cost and overtime penalty cost, and establish constraints at the same time; S2, uses genetic algorithm to optimize and obtain the joint scheduling solution with the lowest total cost under the constraints of S1; S3, perform variable neighborhood local search for the best individual of each generation after genetic operation; S4, perform dynamic perturbation operation; S5, determine whether the maximum number of iterations has been reached. If so, the process ends and a multi-center mixed fleet joint scheduling solution is obtained. Otherwise, repeat S3 to S4.

2. The multi-center mixed fleet joint scheduling method based on the improved genetic variable neighborhood algorithm according to claim 1 is characterized in that: In the joint scheduling model, in the distribution link, the freight volume, distance, delivery time window, and delivery duration of the customer point are known, and vehicles are dispatched by each logistics distribution center to meet the customer's freight needs; all vehicles depart from each logistics distribution center and do not need to return to the original logistics distribution center after the freight task is completed, but can stop at any nearby logistics distribution center; electric vehicles have a minimum power threshold, and when the power is insufficient during the journey, they need to go to the charging station to replenish electricity; vehicles are not allowed to serve customers overtime, otherwise penalty fees will be incurred.

3. The multi-center mixed fleet joint scheduling method based on the improved genetic variable neighborhood algorithm according to claim 2 is characterized in that: In S1, the objective function is to minimize the total cost F consisting of vehicle transportation cost F1, charging cost F2 and overtime penalty cost F3, and establish a joint scheduling model as follows: minF=F1+F2+F3 Among them, the vehicle transportation cost F1 is expressed by the following formula: The charging cost F2 is expressed by the following formula: The overtime penalty cost F3 is expressed by the following formula: Among them, f1 and f2 represent the unit transportation costs of electric vehicles and fuel vehicles respectively; d ij represents the distance of arc (i, j); represents a 0-1 variable, which is 1 if the vehicle v passes through the arc (i, j) and 0 if not; represents the charging amount of electric vehicle v at charging station r; f3, f4 represent the unit charging cost and unit time penalty cost respectively; b i represents the time window of customer i; α iv Indicates the time when the vehicle arrives.

4. The multi-center mixed fleet joint scheduling method based on the improved genetic variable neighborhood algorithm according to claim 3 is characterized in that: The constraints set in S1 include: each customer is delivered by a vehicle only once; the load of electric vehicles and fuel vehicles is not greater than their respective capacities; vehicles comply with cargo flow balance constraints; vehicles can stop at any logistics distribution center on the return trip; vehicles have flow balance constraints at customer points; electric vehicles are charged at most once; the number of vehicles at each center is limited; the relationship between the amount of electricity before and after the electric vehicle passes through arc (i, j); the linear relationship between the charge amount and the charging rate; the minimum charge limit constraint for electric vehicles; the amount of electricity after charging of the electric vehicle must be able to meet the remaining travel requirements; the electric vehicle starts with a full charge; the vehicle leaves from point i, and if it passes through arc (i, j), the time to arrive at point j is greater than or equal to the departure time of point i plus the travel time on arc (i, j); the time constraints for the vehicle to arrive at and leave the customer point and the charging station.

5. The multi-center mixed fleet joint scheduling method based on the improved genetic variable neighborhood algorithm according to any one of claims 1 to 4, characterized in that: S2 specifically includes the following sub-steps: S21, set the initial parameters, including the maximum number of iterations, population size, crossover rate, and mutation rate; S22, uses greedy clustering to construct the population to generate a high-quality initial population; S23, taking the inverse of the objective function F (1 / F) as the fitness function and performing fitness calculation; S24, sort the individuals in the population from high to low according to their fitness, and add the preset proportion of individuals with the highest ranking to the offspring population; S25, random order crossover was performed; S26, performs parallel mutation; S27, for EV routes with infeasible power, a backtracking method is used to insert charging stations and repair infeasible solutions to meet actual constraints.

6. The multi-center mixed fleet joint scheduling method based on the improved genetic variable neighborhood algorithm according to claim 5 is characterized in that: S22 includes the following sub-steps: S221 divides all customers into corresponding logistics distribution centers according to the principle of closest distance, and obtains multiple unit clusters after clustering each logistics distribution center; In the S222 unit cluster, a customer is randomly selected and an initial solution path that satisfies the vehicle capacity constraint is constructed according to the greedy criterion; S223, repeat S222 until the population individuals reach the scale.

7. The multi-center mixed fleet joint scheduling method based on the improved genetic variable neighborhood algorithm according to claim 5 is characterized in that: S27 includes the following sub-steps: S271, checking the circuit to determine the position of the breakpoint where the power is insufficient; S272, inserting charging stations with the shortest distances to the front and rear customers in the interval from the breakpoint position in sequence from the back to the front; S273, calculating the incremental cost of the line after each charging station is inserted, and inserting the charging station at the position with the minimum incremental cost to obtain a feasible solution after repair.

8. The multi-center mixed fleet joint scheduling method based on the improved genetic variable neighborhood algorithm according to claim 5 is characterized in that: Design neighborhood structure N in S3 k , which includes 6 different operators, namely N1 Swap: swap the positions of two customers in the current path; N2 2-opt: reverse the middle segment of two customers in the current path; N3 Inter Swao 1-1: select a customer on each of the two paths for swap; N4 1-2-3 insert: select 1, 2, and 3 customers from three paths respectively and remove them, and then reinsert them into the corresponding paths in the order of 1→2, 2→3, and 3→1; N5 Depot Swap 1-1: select a customer on each path in two distribution centers for swap; N6 Delete Smallest Route: delete the path with the least customers and put them back into the remaining path.

9. The multi-center mixed fleet joint scheduling method based on the improved genetic variable neighborhood algorithm according to claim 8 is characterized in that: S3 includes the following sub-steps: S31, input the current solution x, set k = 1; S32, using the neighborhood structure N k When performing local search, the main thread distributes the current solution x to the first k neighborhood structures at the same time, and each neighborhood structure acts as a thread to generate candidate solutions x in the current neighborhood. k And calculate the corresponding target value; S33, the main thread compares the target values ​​of all candidate solutions and selects the best candidate solution x best ; S34, the main thread compares the best candidate solution x best and the current solution x target value to determine whether an improvement is obtained. If so, update the current solution to the optimal candidate solution and return to S31 to continue searching; otherwise, keep the current solution unchanged and set k = k + 1; S35, check whether the neighborhood search termination condition is met. If k≤6, return to S32 and restart the search. Otherwise, end the search process and enter S4.

10. The multi-center mixed fleet joint scheduling method based on the improved genetic variable neighborhood algorithm according to claim 9 is characterized in that: S4 includes the following sub-steps: S41, is the neighborhood structure N k The six operators included are given the same weight so that each operator has the same probability of being selected; S42, during the perturbation process, evaluates the quality improvement effect of the solution before and after the application of each operator; S43, dynamically assign new weights to operators based on the improvement effects, including increasing the weights of operators with outstanding improvement effects and reducing the weights of operators with poor performance. Operators with excellent performance are given priority in subsequent disturbances.

Images