Shared warehouse and oil-electricity hybrid motorcade path optimization method considering battery replacement

Abstract

The invention relates to a shared warehouse and oil-electricity hybrid motorcade path optimization method considering battery replacement. The method comprises the following steps: establishing an order allocation scheme of a hybrid motorcade joint mode according to the freight volume and a distribution group; aiming at the order distribution scheme, constructing a hybrid motorcade path optimization model taking minimization of the total cost of the delivery task as a target function; and solving the hybrid motorcade path optimization model based on a hybrid heuristic algorithm to obtain an optimal path meeting the constraint of the hybrid motorcade path optimization model. According to the method, an order distribution mode is determined according to different customer group freight volumes and distribution groups, a cooperative logistics mode of a traditional warehouse and a third-party electric vehicle battery sharing warehouse is constructed, the e-commerce material distribution efficiency is improved while the e-commerce urban logistics operation cost is reduced, the e-commerce product distribution cost and time are effectively reduced, and the system is suitable for large-scale popularization and application. And the e-commerce supply chain is enabled from the dimension of urban logistics, so that the business benefit is improved.

Description

Technical Field

[0001] The present application relates to the field of logistics and transportation technology, and in particular to a method for optimizing the routes of shared warehouses and hybrid electric vehicle fleets taking battery replacement into consideration. Background Art

[0002] In the context of multi-point distribution in the community and special materials with capacity classification, the complexity and particularity of the current vehicle distribution model are constantly highlighted. Due to the impact of the boom in e-commerce and environmental protection and the changes in the consumption structure of urban residents, the traditional vehicle distribution operation mechanism based on experience has the problems of low efficiency and insufficient cost optimization. It is of great significance to build a distribution model of mixed fleet operation of electric vehicle rental services and consider the sharing mechanism of coordinated distribution of electric vehicles and fuel vehicles.

[0003] Electric vehicles have lower purchase costs, operating costs and ease of operation, but compared to electric vehicles, fuel vehicles have advantages such as longer driving distance, larger cargo capacity and more convenient refueling. Therefore, how to reasonably and comprehensively utilize the advantages of electric vehicles and fuel vehicles and optimize the heterogeneous vehicle delivery route planning is the key to successfully addressing the challenges of large-scale, high complexity and limited battery replacement strategies faced by electric vehicle collaborative delivery in the e-commerce platform environment. Summary of the invention

[0004] In order to solve or partially solve the problems existing in the related technologies, the present application provides a shared warehouse and hybrid fleet path optimization method considering battery replacement, formulates a sharing mechanism considering the coordinated distribution of electric vehicles and fuel vehicles, and allocates orders according to the different freight volumes of customer groups and distribution groups.

[0005] The first aspect of the present application provides a method for optimizing the path of a shared warehouse and a hybrid vehicle fleet considering battery swapping, comprising the following steps:

[0006] First, an order allocation plan for a mixed fleet joint model is established based on freight volume and delivery groups;

[0007] Secondly, a hybrid fleet route optimization model is constructed for the order allocation scheme with the objective function of minimizing the total cost of the delivery task;

[0008] Finally, the mixed fleet path optimization model is solved based on the hybrid heuristic algorithm, and the optimal path that meets the constraints of the mixed fleet path optimization model is obtained.

[0009] Among them, the constraints of the mixed fleet path optimization model include: inlet and outlet flow constraints, input quantity constraints, load constraints, energy consumption constraints and time constraints.

[0010] Among them, the solution of the mixed fleet path optimization model based on the hybrid heuristic algorithm includes:

[0011] Generate an initial acceptable solution based on the perturbation method and vehicle type priority selection strategy initialized by the algorithm;

[0012] Based on the rich neighborhood structure and update strategy, the search range of the initial acceptable solution is expanded to generate a candidate subset that meets the model constraints;

[0013] Search and optimize the candidate subsets and optimize the optimal solution;

[0014] After determining that the termination calculation conditions are met, the optimal path solution is obtained.

[0015] The generation of an initial acceptable solution includes the following steps:

[0016] Step 1: Based on the load limits of different vehicle models and the time window requirements of the first customer point, select a distribution center that meets these conditions and does not violate the minimum battery power requirements as the starting point;

[0017] Step 2: Consider all customer points as unserved points, determine whether the next customer point meets the service time window, vehicle load, and whether the remaining energy supply meets the model constraints, establish a set of customer points that meet the requirements for each vehicle type, and insert the customer node into the path, repeat step 2, and the customer points that do not meet the remaining power requirements of the next task proceed to step 3 for battery replacement, otherwise proceed to step 4; input the disturbance amount to process the infeasible solution;

[0018] Step 3: After delivering the logistics order of the current customer point, if the power level cannot guarantee that the electric vehicle can reach the next customer point for delivery and find the nearest battery swap center for battery swap, repeat step 2, otherwise go to step 4.

[0019] Step 4: Construct a path based on the basic insertion operation, select an unvisited customer from the customer nodes that meet the constraints, and insert it into the path with the least increase in the target total cost and its optimal position;

[0020] Step 5: Send a new vehicle to repeat steps 1 to 4 to continue constructing a path that satisfies the constraints involved in the model until all customer points are divided and served by logistics delivery vehicles.

[0021] Among them, the infeasible solutions for input disturbance processing include:

[0022] The sum of disturbances is limited by multiple inequality constraints;

[0023] Determine whether there is an infeasible solution that does not meet the constraints;

[0024] If there is an infeasible solution, a positive disturbance is input into the model, including: adding the sum of the disturbances to the objective function value to avoid falling into the local optimal solution.

[0025] The step of searching and optimizing the candidate subset and optimizing the optimal solution includes:

[0026] After generating the candidate subset;

[0027] Update the candidate subset taboo table and compare the neighborhood solutions;

[0028] Select a better solution to replace the current solution.

[0029] Among them, the total cost of the delivery task includes the fixed cost of fuel vehicles and electric vehicles, travel costs and time penalty costs.

[0030] The shared warehouse and hybrid fleet route optimization method also includes a load energy consumption model, which expresses the relationship between energy and actual load as a linear consumption relationship.

[0031]

[0032] In the formula, Q k Indicates the maximum payload of a fuel vehicle, Q u represents the maximum payload of the electric vehicle, Φ represents the set of fuel vehicles, k∈Φ represents any vehicle in the set, Ψ represents the set of electric vehicles, and u∈Ψ represents any vehicle in the set.

[0033] Among them, according to the different order allocation modes of the freight volume of the customer group and the delivery group, the order allocation plan divides the orders to be delivered into fuel vehicle freight plans with large freight volume and long delivery distance and electric vehicle freight plans with small freight volume and short delivery distance.

[0034] The technical solution provided by this application may have the following beneficial effects:

[0035] The present application provides a method for optimizing the routes of shared warehouses and hybrid electric vehicle fleets that takes battery swapping into consideration, establishes an order allocation model based on the freight volume of customer groups and the differences in distribution groups, and constructs a cooperative logistics model between traditional warehouses and third-party electric vehicle battery shared warehouses. While reducing the operating costs of e-commerce urban logistics, it improves the efficiency of e-commerce material distribution, effectively reduces the cost and time of e-commerce product distribution, and empowers the e-commerce supply chain from the dimension of urban logistics to improve business benefits.

[0036] It should be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] The above and other objects, features and advantages of the present application will become more apparent through a more detailed description of exemplary embodiments of the present application in conjunction with the accompanying drawings, wherein the same reference numerals generally represent the same components in the exemplary embodiments of the present application.

[0038] Figure 1 It is a flowchart of a path optimization method shown in an embodiment of the present application;

[0039] Figure 2 It is a schematic diagram of the solution process of the hybrid heuristic algorithm shown in the embodiment of the present application. DETAILED DESCRIPTION

[0040] The embodiments of the present application will be described in more detail below with reference to the accompanying drawings. Although the embodiments of the present application are shown in the accompanying drawings, it should be understood that the present application can be implemented in various forms and should not be limited by the embodiments described herein. On the contrary, these embodiments are provided to make the present application more thorough and complete, and to fully convey the scope of the present application to those skilled in the art.

[0041] It should be understood that although the terms "first", "second", "third", etc. may be used in this application to describe various information, this information should not be limited to these terms. These terms are only used to distinguish the same type of information from each other. For example, without departing from the scope of this application, the first information may also be referred to as the second information, and similarly, the second information may also be referred to as the first information. Thus, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features. In the description of this application, the meaning of "multiple" is two or more, unless otherwise clearly and specifically defined.

[0042] In the description of the present application, it should be understood that the terms "length", "width", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", etc., indicating the orientation or position relationship, are based on the orientation or position relationship shown in the drawings, and are only for the convenience of describing the present application and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as a limitation on the present application.

[0043] Unless otherwise clearly specified and limited, the terms "installed", "connected", "connected", "fixed" and the like should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium, it can be the internal connection of two elements or the interaction relationship between two elements. For ordinary technicians in this field, the specific meanings of the above terms in this application can be understood according to specific circumstances.

[0044] The technical solution of the embodiments of the present application is described in detail below with reference to the accompanying drawings.

[0045] like Figure 1 The route optimization method for a shared warehouse and a hybrid fleet considering battery swapping shown in the figure includes the following steps:

[0046] S1. Establish an order allocation plan for a mixed fleet joint model based on freight volume and distribution groups.

[0047] Based on the SE-HMDVRP (Shared Electric Vehicle-Multi Depot Vehicle Routing Problem) problem of e-commerce platform trucks (fuel vehicles) and third-party electric vehicles, a directed graph was constructed to describe the distribution network of e-commerce products. In view of the characteristics of B2C e-commerce logistics, a different order allocation mode based on the freight volume of customer groups and distribution groups was proposed. Suppliers divide the orders to be delivered into e-commerce platform truck freight solutions with large freight volume and long delivery distance and third-party shared electric vehicle freight solutions with small freight volume and short delivery distance. Through the order allocation form of distance and capacity classification processing, electric vehicles are used as much as possible for short-distance and small-capacity orders. While reducing the operating costs of e-commerce urban logistics, the efficiency of e-commerce material distribution is improved.

[0048] Establish a mixed fleet of fuel vehicles and electric vehicles for joint delivery, and assign different vehicle models to central warehouses with different functions. E-commerce platform trucks can only return to the central warehouse to pick up goods. After completing the current delivery task, third-party electric vehicles do not need to return to the central warehouse to pick up goods, but can pick up goods at the nearest operation warehouse for delivery. When the vehicle has no energy to carry goods, the community's operation center has abundant fuel or batteries to power the vehicle to ensure that it can normally perform the next stage of delivery tasks. This solution is adopted to reduce the cost and delivery time of e-commerce product delivery, empower the e-commerce supply chain from the dimension of urban logistics, and thus improve business benefits.

[0049] S2. For the order allocation scheme, a mixed fleet path optimization model is constructed with minimizing the total cost of the delivery task as the objective function.

[0050] Model assumptions for the mixed fleet routing optimization model include:

[0051] 1) Ignoring the impact of road traffic conditions, vehicles drive steadily and at a constant speed when performing delivery tasks;

[0052] 2) The energy consumption of a vehicle is related to its current load, and its maximum energy is limited by the capacity of the fuel tank or electric vehicle battery;

[0053] 3) When the fuel level of the e-commerce platform truck is sufficient to support the entire delivery, and the power of the third-party shared electric vehicle is insufficient to support the next task, it will go to the shared battery replacement warehouse for battery replacement;

[0054] 4) A customer can only use one delivery plan to deliver his goods, and the same customer's order cannot be split and delivered;

[0055] 5) To simplify the model, it is assumed that the relationship of energy consumption is linear and there is no overload when starting from the node.

[0056] In order to reduce the complexity of solving the energy consumption of HEO-HMDVRP during the battery replacement process and the difficulty of process quantity measurement, the load energy consumption model is introduced to obtain the relationship between energy consumption and vehicle load. The energy consumption model is simplified to a linear relationship:

[0057]

[0058]

[0059] The relationship between energy and actual load is simplified to a linear consumption relationship, ignoring the impact of traffic congestion and weather conditions.

[0060] In order to save the total distribution cost of enterprises in the mixed fleet joint mode, a single-objective MILP model mixed fleet path optimization model is constructed. Its specific objectives and constraints are as follows:

[0061] Objective function:

[0062]

[0063] Constraints: Inbound and outbound traffic constraints:

[0064] Input quantity constraints:

[0065] Load capacity constraints:

[0066] Energy Consumption Constraints:

[0067] Time constraints:

[0068]

[0069] Among them, objective function (3) represents minimizing the total cost of the delivery task, including the fixed cost, travel cost and time penalty cost of fuel vehicles and electric vehicles. Constraints (4) and (5) ensure that fuel vehicles return to the warehouse after leaving the warehouse. Constraints (6) and (7) indicate that electric vehicles visit each customer point after leaving the warehouse. Constraints (8) and (9) limit vehicle flow, and fuel vehicles and electric vehicles visit a customer point at most once. Constraints (10) and (11) ensure that each fuel vehicle and electric vehicle leaves the customer point immediately after visiting a customer point, that is, the flow protection of the vehicle reaching a customer point. Constraint (12) indicates that for a certain customer point, there is at most one type of vehicle that arrives at the point to perform the delivery task.

[0070] Constraints (13) and (14) indicate that the number of fuel vehicles and electric vehicles used in delivery tasks cannot exceed the total number of vehicles. Constraints (15) and (16) limit the number of vehicles used in the same customer point to only one possibility.

[0071] Constraints (17) and (18) indicate that the total weight of the goods delivered by fuel vehicles and electric vehicles that provide services to customer points cannot exceed their maximum payload, that is, the vehicle cannot be overweight when performing a delivery task. Constraints (19) and (20) indicate that the remaining payload of fuel vehicles and electric vehicles leaving a node must not be less than zero, that is, the remaining payload must support the completion of this task. Constraints (21) and (22) indicate that the remaining payload of a fuel vehicle entering a node is equal to the remaining payload of the vehicle leaving the node plus the weight of the goods delivered to the node. Constraints (23) and (24) indicate that the remaining payload of an electric vehicle entering a node is equal to the remaining payload of the vehicle leaving the node plus the weight of the goods delivered to the node.

[0072] Constraint (25) indicates that the fuel energy of the e-commerce platform fuel car at the starting warehouse is equal to the maximum fuel supply energy of the e-commerce platform fuel car. Constraint (26) indicates that the battery energy of the third-party electric car at the starting warehouse is equal to the maximum battery supply energy of the third-party electric car. Constraint (27) indicates that the fuel energy of the fuel car leaving a node is equal to the fuel energy leaving the previous node minus the fuel consumption of the goods between the two points. Constraint (28) indicates that the battery energy of the electric car leaving a node is equal to the battery energy leaving the previous node minus the electric energy consumption of the goods between the two points. Constraints (29) and (30) indicate that the energy of the fuel car and the electric car leaving a node is equal to the energy entering the node, that is, no energy is consumed when unloading at the node. Constraints (31) and (32) indicate that the energy of the fuel car or the electric car leaving a node is not less than 0, that is, when the remaining energy is insufficient, energy replenishment (refueling or charging) is required.

[0073] Constraints (33) and (34) indicate that the time for a fuel vehicle to leave and enter the warehouse is 0. Constraints (34) and (35) indicate that the time for a fuel vehicle and an electric vehicle to leave a node is equal to the time they enter the node plus the time they stay at the node. Constraints (36) and (37) indicate that the time for a fuel vehicle and an electric vehicle to enter a node is equal to the time they leave the previous node plus the time the vehicle travels between the two points. Constraints (38) and (39) indicate that the time for a fuel vehicle and an electric vehicle to refuel and change points does not exceed the time the vehicle stays at the point.

[0074] The problem is converted into the data representation required by the mathematical model through encoding and decoding:

[0075] According to the decision variables set in the MILP model This three-dimensional variable can represent a digital sequence scheme, describing the distribution scheme of the discrete optimization problem proposed in this patent as a natural number sequence from different sets. Among them, the node set {1,2,3,4,5} represents warehouses in different locations; the customer point set is {6,7,...,71,72}. The corresponding relationship of the path encoding method obtained from the model is as follows: It means that the e-commerce platform truck k1 starts from warehouse 1, visits customer point 7, customer point 13, customer point 18 in sequence, and finally returns to warehouse 1; It means that the third-party electric vehicle starts from the distribution center 2 and visits the customer point 8 in sequence. Due to insufficient power, it goes to the customer point 17 with battery replacement service function, then goes to the customer point 6 for delivery, and finally returns to the warehouse 2.

[0076] S3. Solve the mixed fleet path optimization model based on the hybrid heuristic algorithm to obtain the optimal path that meets the constraints of the mixed fleet path optimization model.

[0077] Considering the complexity of MDVRP, which is an NP-hard problem, most researchers have proposed using modern heuristic algorithms as a solution to MDVRP. Tabu Search (TS) and Simulated Annealing (SA) have been proven to be effective strategies for solving MDVRP of mixed fleets. TS is an extension of local neighborhood search. It sets a taboo table to taboo some operations that have been experienced to guide the next search direction to prevent falling into the local optimum. SA has certain randomness and probability compared to GA. During the search process, it may accept inferior solutions with a certain probability to avoid falling into the local optimum. Based on this, a specific feasible solution is proposed to deal with the complexity of the proposed model, which is due to the semi-open path of the problem and the need for electric vehicles to replace batteries. Combining the principles of the TS-RN algorithm (Necati A. et al., 2011) and the SA algorithm, a hybrid heuristic algorithm TS-RN&SA is developed to solve the heuristic program of the problem proposed in this study.

[0078] Solving the mixed fleet path optimization model based on the hybrid heuristic algorithm includes the following steps:

[0079] S31. Generate an initial acceptable solution based on the disturbance method initialized by the algorithm and the vehicle model priority selection strategy.

[0080] In order to obtain the optimal solution in solving large-scale optimization problems, the simulated annealing algorithm often requires a high initial temperature and a sufficient number of sampling times, which makes the optimization time of the algorithm too long. The improved simulated annealing algorithm's quality requirements for the initial solution lead to the random generation of the initial solution, while giving priority to the use of electric vehicle models with low fixed costs and distribution costs for distribution, generating low-cost distribution routes, taking the minimum increase in the objective function, that is, the total cost, as the acceptance criterion, and repeatedly comparing to obtain feasible solutions that meet the requirements of time constraints, capacity constraints, energy consumption constraints, etc. involved in this study, which is conducive to improving the quality of the initial solution and avoiding falling into the local optimal situation.

[0081] The specific steps include:

[0082] S311. Based on the load limits of different vehicle models and the time window requirements of the first customer point, a distribution center that meets these conditions and does not violate the minimum battery power requirements is selected as the starting point. If the current vehicle model cannot meet the above conditions, the next vehicle model is selected and the process is repeated.

[0083] S312. Treat all customer points as unserved points, determine whether the next customer point meets the service time window, vehicle load, and whether the remaining energy supply meets the model constraints, establish a set of customer points that meet the requirements for each vehicle type, and insert the customer node into the path, repeat S312, and the customer points that do not meet the remaining power requirements of the next task proceed to S313 for battery replacement, otherwise proceed to S314.

[0084] In order to obtain the global optimal solution in the search space, some infeasible solutions close to feasible solutions or optimal solutions should be allowed to exist in the initialization stage, and in the later stage, infeasible solutions should be gradually eliminated, and at the end of the algorithm, all solutions in the non-inferior solution set should be ensured to be in the feasible solution area. Therefore, this study uses the method of inputting disturbance to deal with infeasible solutions:

[0085]

[0086] Where T(X) represents the sum of the disturbances consisting of n inequality constraints. When the constraints are met, no disturbance is added to the model; otherwise, a positive disturbance is input into the model. First, the total restricted disturbance is calculated; then, the T(X) value is added to the objective function value, so that the search space for the solution is increased and the local optimal solution is avoided.

[0087] S313. After delivering the logistics order of the current customer point, if the power supply cannot guarantee that the electric vehicle can reach the next customer point for delivery and find the nearest battery exchange center for battery exchange, then repeat S312, otherwise go to S314.

[0088] S314: Construct a path based on the basic insertion operation, select an unvisited customer from the customer nodes that meet the constraints, and insert it into the path with the least increase in the target total cost and its optimal position. After all customer path allocations are completed in this stage, the distribution center is selected based on the distance according to the relationship between load and energy consumption, or the distribution center is returned to the selectable center after the battery replacement is completed.

[0089] S315. Send a new vehicle and repeat S311-S314 to continue constructing a path that satisfies the constraints involved in the model until all customer points are divided and served by the logistics delivery vehicles.

[0090] S32. Based on the rich neighborhood structure and update strategy, the search range of the initial acceptable solutions is expanded to generate a candidate subset that meets the model constraints.

[0091] S33, searching and optimizing the candidate subsets and optimizing the optimal solution;

[0092] S34: After determining that the termination calculation conditions are met, an optimal path solution is obtained.

[0093] The rich neighborhood structure and update strategy are used to expand the search scope, form a new candidate subset that meets the model constraints, update the taboo table and compare the neighborhood solutions, select a better solution to replace the current solution, determine whether the conditions for terminating the calculation are met, and obtain the optimal solution.

[0094] In order to optimize the quality of the solution and the calculation speed, the TS-RN algorithm optimizes the search domain and the current solution selection of the original algorithm in terms of neighborhood structure and related Tabu List update strategy, increasing the diversity of the search in the form of disturbance while improving the quality of the solution. In addition, since the taboo search algorithm requires a high-quality initial solution, the above initialization optimization method is used based on the simulated annealing algorithm to generate a high-quality acceptable initial solution, and then the search optimization is performed according to the Improved-TS to optimize the acceptance process of the optimal solution. The specific process is as follows: Figure 2 shown.

[0095] The TS-RN algorithm uses taboo conditions to perform exploration of different regions of the solution space. For this purpose, each time a solution is updated, a taboo condition is created that is associated with a certain neighborhood structure. It is forbidden to ignore the taboo condition unless it immediately improves the best feasible goal, achieving Z. best Search for.

[0096] Model symbol description:

[0097]

[0098]

[0099]

[0100]

[0101] Finally, it should be noted that, in this article, relationships such as first and second, etc. are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms include, include or any other variations are intended to cover non-exclusive inclusion, so that a process, method, article or device that includes a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0102] 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, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0103] The embodiments of the present application have been described above, and the above description is exemplary, not exhaustive, and is not limited to the disclosed embodiments. Many modifications and changes will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the described embodiments. The selection of terms used herein is intended to best explain the principles of the embodiments, practical applications, or improvements to the technology in the market, or to enable other persons of ordinary skill in the art to understand the embodiments disclosed herein.

Claims

1. A method for optimizing the route of a shared warehouse and a hybrid fleet of oil and electricity considering battery swapping, characterized in that: The following steps are involved: Establish order allocation schemes for mixed fleet joint models based on freight volume and delivery groups; According to the order allocation scheme, a mixed fleet path optimization model is constructed with minimizing the total cost of the delivery task as the objective function; The mixed fleet path optimization model is solved based on a hybrid heuristic algorithm to obtain an optimal path that meets the constraints of the mixed fleet path optimization model.

2. The shared warehouse and hybrid fleet route optimization method according to claim 1, characterized in that: The constraints of the mixed fleet path optimization model include: inflow and outflow flow constraints, input quantity constraints, load constraints, energy consumption constraints and time constraints.

3. The shared warehouse and hybrid fleet route optimization method according to claim 1, characterized in that: The method of solving the mixed fleet path optimization model based on the hybrid heuristic algorithm includes: Generate an initial acceptable solution based on the perturbation method and vehicle type priority selection strategy initialized by the algorithm; Expanding the search range of the initial acceptable solution based on the rich neighborhood structure and the updating strategy to generate a candidate subset that meets the model constraints; Searching and optimizing the candidate subset and optimizing the optimal solution; After determining that the termination calculation conditions are met, the optimal path solution is obtained.

4. The shared warehouse and hybrid fleet route optimization method according to claim 3 is characterized in that: The generating of the initial acceptable solution comprises the following steps: Step 1: Based on the load limits of different vehicle models and the time window requirements of the first customer point, select a distribution center that meets these conditions and does not violate the minimum battery power requirements as the starting point; Step 2: Consider all customer points as unserved points, determine whether the next customer point meets the service time window, vehicle load, and whether the remaining energy supply meets the model constraints, establish a set of customer points that meet the requirements for each vehicle type, and insert the customer node into the path, repeat step 2, and the customer points that do not meet the remaining power requirements of the next task proceed to step 3 for battery replacement, otherwise proceed to step 4; input the disturbance amount to process the infeasible solution; Step 3: After delivering the logistics order of the current customer point, if the power level cannot guarantee that the electric vehicle can reach the next customer point for delivery and find the nearest battery swap center for battery swap, repeat step 2, otherwise go to step 4. Step 4: Construct a path based on the basic insertion operation, select an unvisited customer from the customer nodes that meet the constraints, and insert it into the path with the least increase in the target total cost and its optimal position; Step 5: Send a new vehicle to repeat steps 1 to 4 to continue constructing a path that satisfies the constraints involved in the model until all customer points are divided and served by logistics delivery vehicles.

5. The shared warehouse and hybrid fleet route optimization method according to claim 4 is characterized in that: The infeasible solution of the input disturbance quantity processing includes: The sum of disturbances is limited by multiple inequality constraints; Determine whether there is an infeasible solution that does not meet the constraints; If an infeasible solution exists, a positive disturbance is input into the model, including: adding the sum of the disturbance amount to the objective function value to avoid falling into a local optimal solution.

6. The shared warehouse and hybrid fleet route optimization method according to claim 1, characterized in that: The step of searching and optimizing the candidate subset and optimizing the optimal solution includes: After generating the candidate subset; Updating the taboo table of the candidate subset and comparing neighborhood solutions; Select a better solution to replace the current solution.

7. The shared warehouse and hybrid fleet route optimization method according to claim 1, characterized in that: The total cost of the delivery task includes the fixed cost of fuel vehicles and electric vehicles, travel costs and time penalty costs.

8. The shared warehouse and hybrid fleet route optimization method according to claim 1, characterized in that: The shared warehouse and hybrid fleet route optimization method further includes a load energy consumption model, wherein the load energy consumption model represents the relationship between energy and actual load as a linear consumption relationship. In the formula, Q k Indicates the maximum payload of a fuel vehicle, Q u represents the maximum payload of the electric vehicle, Φ represents the set of fuel vehicles, k∈Φ represents any vehicle in the set, Ψ represents the set of electric vehicles, and u∈Ψ represents any vehicle in the set.

9. The shared warehouse and hybrid fleet route optimization method according to claim 1, characterized in that: According to the different order allocation modes of the freight volume of the customer group and the delivery group, the order allocation plan divides the orders to be delivered into a fuel vehicle freight plan with a large freight volume and a long delivery distance and an electric vehicle freight plan with a small freight volume and a short delivery distance.

Images