Three-level urban and rural joint distribution network path planning method based on adaptive hybrid algorithm

Abstract

The invention discloses a three-level urban and rural joint distribution network path planning method based on an adaptive hybrid algorithm, and particularly relates to a distribution path planning method which takes a time window, vehicle capacity and path coherence as constraint conditions and combines an adaptive firework-quantum genetic hybrid algorithm for optimization. The method comprises the following steps: firstly, acquiring basic information such as coordinates, demand quantity, time windows and the like of an urban common distribution center, county and rural transfer stations and rural end self-pickup points, and initializing vehicle resources; then, a mathematical model with the purpose of minimizing the total distribution cost is constructed based on common distribution, wherein the model comprises fixed transportation cost, variable transportation cost, time penalty cost and the like; and solving the model by adopting a self-adaptive fireworks-quantum genetic hybrid algorithm, which comprises the steps of generating an initial solution by quantum bit coding, optimizing the initial solution by an insertion algorithm, carrying out global search on a fireworks algorithm and carrying out local optimization on a quantum genetic algorithm. In the optimization process, spark disturbance operation, quantum bit revolving door dynamic adjustment and crossover mutation operator are combined to improve the diversity and convergence speed of solutions, and finally the optimal distribution path meeting the time window constraint is output. The method is suitable for path planning of a complex three-level urban and rural distribution network, the transportation cost can be remarkably reduced, the distribution efficiency is improved, and urban and rural logistics resource allocation is optimized.

Description

Technical Field

[0001] The present invention relates to logistics distribution optimization technology, and in particular to a method for path planning of a three-level urban-rural joint distribution network based on an adaptive hybrid algorithm. Background Art

[0002] In recent years, with the rapid development of e-commerce and logistics technology, the scale of my country's rural e-commerce market has expanded rapidly. However, the rural logistics system, as the pillar of rural e-commerce, still faces prominent problems such as unbalanced resource allocation, high logistics costs, and low transportation efficiency, especially in urban-rural fringe areas or low distribution density areas.

[0003] To address the above problems, a three-level urban-rural distribution network based on joint distribution has gradually become an innovative modern logistics model. By sharing the resources, vehicles and information platforms of urban joint distribution centers, county and township transfer stations and rural terminal pick-up points, joint distribution can effectively integrate the resources of distribution nodes at all levels from urban to rural areas, reduce transportation costs and improve distribution efficiency. However, there are many studies on the joint distribution vehicle routing problem, but there is a lack of optimization models specifically for urban-rural three-level distribution. Most studies plan the two-stage distribution path in stages, without considering the impact of the time-varying demand of the terminal node on the existing vehicle routing planning. In addition, traditional path planning algorithms, such as genetic algorithms and ant colony algorithms, have problems of high computational complexity and unstable results, and are difficult to adapt to the complex three-level distribution network. In the practice of applying joint distribution to the three-level urban-rural distribution network, it still faces the challenges of complex path planning systems, optimization of distribution stages with different requirements, and multi-objective optimization requirements such as vehicles and time windows. In view of this, it is urgent to propose a path planning method for the three-level urban-rural joint distribution network based on an adaptive hybrid algorithm, and use an improved heuristic algorithm of an adaptive fireworks-quantum genetic hybrid algorithm to solve it, so as to improve the path planning efficiency and optimization effect. Summary of the invention

[0004] In view of the technical deficiencies in the research of this field, the present invention proposes a method for path planning of a three-level urban-rural joint distribution network based on an adaptive hybrid algorithm. The present invention considers constraints such as distribution path, service time window, and maximum vehicle load, and establishes a vehicle distribution path determination model with the minimum total distribution cost including fixed transportation cost, variable transportation cost, and time penalty cost as the optimization goal. An improved heuristic algorithm of an adaptive fireworks-quantum genetic hybrid algorithm is used to solve the model to improve the efficiency of path planning and optimization effect.

[0005] The present invention is a method for path planning of a three-level urban-rural joint distribution network based on an adaptive hybrid algorithm, characterized in that:

[0006] To achieve the above objectives, the present invention is implemented through the following technical solutions:

[0007] A three-level urban-rural distribution network system based on joint distribution, comprising:

[0008] The system includes a three-level distribution network. First, in order to solve the problems of unbalanced resource allocation, high logistics costs, and low transportation efficiency, the urban and rural distribution tasks and other resources of multiple logistics companies are integrated, and vehicle resources, customer resources, and warehousing resources are jointly built and shared to build a common distribution platform. Secondly, the urban and rural distribution network includes a first-level urban common distribution center responsible for integrating cargo resources within the city and distributing the cargo to county and township transfer stations; several second-level township and county transfer stations as intermediate hubs, responsible for receiving cargo from urban distribution centers and further distributing it to rural terminal nodes; and several third-level rural terminal nodes provide self-pickup services to solve the problem of the last mile of rural distribution. The goods required by customers are transported by large vehicles from the urban common distribution center to various county and township transfer stations, and then small vehicles transport the packages from the county and township transfer stations to various rural terminal self-pickup points. In the three-level urban and rural distribution network, the township and county transfer stations play the role of linking the upper-level urban common distribution centers and the lower-level farm terminal pick-up points. In the upper-level distribution network, goods are transported from the urban common distribution centers to the township and county transfer stations by large vehicles. In the lower-level distribution network, goods are transported from the township and county transfer stations to the various rural terminal pick-up points by small vehicles. In the upper-level distribution network, the demand for township and county transfer stations may be greater than the maximum capacity of the distribution vehicles. Therefore, the upper-level distribution network belongs to demand-separable distribution, and the lower-level distribution network belongs to demand-inseparable distribution.

[0009] A method for path planning of a three-level urban-rural joint distribution network based on an adaptive hybrid algorithm; comprising:

[0010] S1. Obtain vehicle resources, status and delivery task information of the common delivery platform;

[0011] S2. constructing a mathematical model with the goal of minimizing the total delivery cost based on the vehicle resources and status and delivery task information;

[0012] S3. An improved heuristic algorithm of an adaptive fireworks-quantum genetic hybrid algorithm is used to solve the mathematical model and obtain the optimal vehicle path of the three-level urban-rural distribution network model, including:

[0013] In the first stage, an initial solution is generated by quantum bit encoding to represent the delivery path. Using quantum bits to represent the solution, the position of each delivery node is initialized in the form of a quantum state, multiple potential paths are generated through the superposition state of quantum bits, and the initial solution is selected through quantum measurement.

[0014] In the second stage, in order to improve the quality of the initial solution, the insertion algorithm is used to optimize the initial solution. The insertion algorithm optimizes the initially generated solution by selecting the insertion position with the minimum increment, that is, when inserting an unvisited node into an existing path, the position that minimizes the total cost of the path is selected.

[0015] In the third stage, the fireworks algorithm is used for optimization. The fireworks algorithm works after the initial solution is generated, and uses sparks to explore better solutions in the solution space. Each spark represents a solution, and the number of sparks and the explosion radius are adaptive and gradually adjusted during the iteration process. During the optimization process of the fireworks algorithm, each spark performs a local search for the current solution. By randomly perturbing the spark position, the algorithm can find a better solution in the local solution space.

[0016] In the fourth stage, the quantum genetic algorithm is used to continue to update the optimal solution. In the quantum genetic algorithm, quantum bit encoding is used to represent each individual, that is, each solution. Each individual represents its path through the state of the quantum bit, and the quantum bit encoding can effectively represent the decision space. In each iteration, new solutions are continuously generated through quantum bit encoding and quantum genetic operations, optimizing the objective function and improving path planning.

[0017] In the fifth stage, non-dominated sorting and selection, it is determined whether the termination conditions are met and the optimal solution is output.

[0018] According to a preferred embodiment of the present invention, the step S2 calculates the criterion parameters of vehicle distribution based on the basic information and establishes a mathematical model with the goal of minimizing the total distribution cost; including:

[0019] The criteria parameters include the transportation cost of the upper-level distribution vehicle, the transportation cost of the lower-level distribution vehicle, and the time penalty cost;

[0020] The calculation formula of the upper distribution vehicle transportation cost C1 is:

[0021]

[0022] Among them, c1 represents the unit distance transportation cost of the upper network distribution vehicle, d ij represents the distance from node i to node j, r iju It is a 0-1 variable, indicating whether vehicle u passes through nodes i and J, D represents the set of urban common distribution centers, S represents the set of township and county transfer stations, and U represents the set of upper-level distribution vehicles.

[0023] The calculation formula of the lower-level distribution vehicle transportation cost C2 is:

[0024]

[0025] Among them, c2 represents the unit distance transportation cost of the lower-level network distribution vehicle, dij represents the distance from node i to node j, x ijk It is a 0-1 variable, indicating whether vehicle k passes through nodes i and j, S represents the set of township and county transfer stations, P represents the set of rural terminal pick-up points, and K represents the set of lower-level distribution vehicles.

[0026] The calculation formula of the time penalty cost c3 is:

[0027]

[0028] Where c3 represents the unit cost coefficient of delay penalty, T i represents the latest arrival time expected by customer i (i.e., the upper limit of the time window), t i represents the time point when the vehicle arrives at the rural terminal self-pickup point i, and P represents the set of rural terminal self-pickup points;

[0029] According to the above criteria parameters, the objective function of minimizing the total distribution cost is established, which is expressed as:

[0030]

[0031]

[0032] Further preferably, the delivery time of the terminal node, the coherence of the delivery path, the vehicle capacity, and the node demand satisfaction are used as the constraint conditions of the objective function of the minimum total delivery cost, and a three-level urban-rural distribution network model based on joint distribution is constructed, including:

[0033]

[0034] In the formula, z su is the transportation volume when the upper-level distribution vehicle u visits the transfer station s, Q1 is the maximum capacity constraint of the upper-level distribution network vehicles;

[0035]

[0036] In the formula, q i is the demand of the terminal rural pick-up point i;

[0037]

[0038] In the formula,

[0039]

[0040] In the formula,

[0041]

[0042] In the formula, e s is the demand for transfer stations in townships and counties;

[0043]

[0044]

[0045] Where M is a sufficiently large real number;

[0046] T iu +d is / v 1is +M×(1-r isu )≥T su i∈D∪S,s∈S,u∈U (15)

[0047] T iu +d is / v 1is -M×(1-r isu )≤T su i∈D∪S,s∈S,u∈U (16)

[0048] Where, T iu is the time it takes for the upper-level delivery vehicle u to arrive at node i, d is is the distance between node i and node s, v 1is is the speed of the upper-level delivery vehicle u from node i to node s, T su is the time when the upper-level delivery vehicle u arrives at the township or county transfer station s;

[0049] T ik +d ij / v 2ij +max(T su )+M×(1-x ijk )≥T jk i∈S∪P,s∈S,j∈P,k∈K (17)

[0050] T ik +d ij / v 2ij +max(T su )-M×(1-x ijk )≤T jk i∈S∪P,s∈S,j∈P,k∈K (18)

[0051] Where, T ik is the time it takes for the lower-level delivery vehicle k to arrive at node i, d ij is the distance between node i and node j, v 2ij is the speed of the lower-level delivery vehicle k from node i to node j, T jkis the time it takes for the lower-level delivery vehicle k to arrive at the rural terminal pick-up point j;

[0052] t i =max(t ik ) (19)

[0053] Where, t i The time it takes for the lower-level delivery vehicle k to arrive at the rural terminal pickup point i;

[0054] in,

[0055] Formula (5) represents the maximum load constraint of vehicles in the upper distribution network;

[0056] Formula (6) represents the maximum load constraint of the vehicles in the lower-level distribution network, and Q2 is the maximum capacity constraint of the vehicles in the lower-level distribution network;

[0057] Equations (7) and (8) indicate that in the upper distribution network, each node j must have only one vehicle reaching it from other nodes, and each node i must have only one vehicle departing from it, ensuring the integrity and continuity of the path;

[0058] Formula (9) and Formula (10) indicate that each terminal rural pick-up point must be visited by a vehicle and can only be visited and served once and only one vehicle can depart from it;

[0059] Formula (11) indicates that the demand distribution of the township and county transfer stations is completed;

[0060] Formula (12) indicates that each vehicle in the upper distribution network starts from the urban common distribution center, passes through several township and county transfer stations, and then returns to the urban common distribution center;

[0061] Formula (13) indicates that each vehicle in the lower-level distribution network starts from a township or county transfer station, passes through several terminal rural pick-up points, and then returns to the township or county transfer station where it started;

[0062] Formula (14) represents the cargo demand of the township and county transfer stations;

[0063] Formula (15)-Formula (19) represents the time when the goods arrive at the demand point;

[0064] Preferably, according to the present invention, the improved heuristic algorithm based on the adaptive fireworks-quantum genetic hybrid algorithm described in step S3 solves the optimal path optimization model to determine the optimal vehicle delivery path, including:

[0065] a. Encode quantum bits for each township and county transfer station and terminal rural node. Each node or terminal demand point is represented by one quantum bit:

[0066] |ψ i >=α i |0>+βi |1> (20)

[0067] Among them, |α i | 2 and |β i | 2 They represent the probability of node selection state, respectively. |0> means the node is not selected, and |1> means the node is selected. A path is represented by a combination of multiple quantum bits.

[0068] b. Initialize the population and randomly set the initial amplitude α of each quantum bit i and β i , guarantee |α i | 2 +|β i | 2 =1, and then the quantum state |ψ i >Decode to a specific selection value of 0 or 1, combine the selected nodes into a path, assign the unselected nodes to other vehicles, randomly generate N candidate initial solutions based on the superposition state of the quantum bit, check the legitimacy of each path after decoding, and repair the solutions that do not meet the constraints;

[0069] c. Optimize the initial solution. Use the insertion algorithm to optimize the initial solution, reduce the path cost and improve the feasibility of the solution. For the path of each vehicle in the initial solution, traverse the unvisited nodes in turn, insert the unvisited node i into all possible positions in the path, and calculate the path cost increment of each insertion scheme:

[0070] ΔC=C new -C original (twenty one)

[0071] Among them, C new represents the path cost after inserting the node, C original Represents the path cost before inserting the node. In order to increase the diversity of the initial population, the present invention selects the position with the smallest cost increment to insert node i and updates the path. After all nodes are inserted, the optimized path is obtained. Check whether the optimized path meets the constraints of capacity and time window. If not, adjust the node order or reallocate the node to other vehicle paths to ensure that the optimized initial solution is feasible.

[0072] d. Use the fireworks algorithm to perform dynamic global search. First, calculate the number of sparks N at the center of each firework using formula (22): s , where N max ,N minRespectively represent the maximum and minimum number of sparks, F(x) represents the fitness value of the current solution x, reflecting the quality of the solution. Then, for each individual vehicle, an explosion operation is used to generate sparks to determine the spark position. The spark generation position is determined by the disturbance range R, which is determined by formula (23), where R represents the explosion radius, and Δx represents the random disturbance factor, which is randomly generated in the range [-1,1]. Calculate each spark solution x s The fitness value F(x s ), for the fireworks center with better fitness, a smaller explosion radius is set to generate a fine spark solution; for the fireworks center with poor fitness, a larger explosion radius is set to expand the search range;

[0073]

[0074] x s =x+R·Δx,Δx∈[-1,1] (23)

[0075] Preferably, the node or path of the current path solution x is disturbed to generate N s Spark solutions, design the following three perturbation operators:

[0076] Node exchange operator: randomly obtain two nodes in the path, exchange their positions, and generate a new path solution;

[0077] Node insertion operator: Using equation (24), move a node i in the path to another position j;

[0078] x′=Insert(x,i,j) (24)

[0079] Path splitting operator: splits a path into two parts and redistributes them to different vehicles.

[0080] Preferably, if the spark solution does not satisfy the constraints, its fitness value is reduced by a penalty function:

[0081] F penalized =F(x s )+λ·violation_penalty

[0082] Preferably, among the newly generated spark solutions, the spark solution with the highest fitness is selected as the new firework center, and the fitness F(x new ) and the global optimal solution F best , if the current solution is better, update the global optimal solution:

[0083] F best =min(F best ,F(x new )) (25)

[0084] Preferably, retain N optimal firework centers as the initial solution for the next round of iteration;

[0085] e. Use the quantum genetic algorithm to locally optimize the solution of the fireworks algorithm. First, use equation (26) to dynamically adjust the rotation angle θ according to the fitness to perform quantum revolving door to update the population:

[0086]

[0087] Preferably, the rotation amplitude is dynamically adjusted according to the fitness value of the solution, so as to locally strengthen the high-quality solution and increase the randomness of the low-quality solution, so as to avoid falling into the local optimum;

[0088]

[0089] Preferably, the following two operators are designed:

[0090] Crossover operator: randomly select two path solutions from the population, exchange some of their nodes, and generate a new path solution that inherits the excellent characteristics of the parent generation:

[0091] x new =Crossover(x1,x2) (29)

[0092] Mutation operator: randomly adjusts the order of nodes in the path, randomly selects two nodes in the path to swap positions, and randomly inserts or deletes a node;

[0093] Preferably, the fitness values ​​of the solutions after crossover and mutation are calculated, solutions with higher fitness values ​​are retained, solutions with lower fitness values ​​are eliminated, the newly generated solutions are merged with the current population, the N solutions with the highest fitness values ​​are selected as the new population and the global optimal solution is updated;

[0094] f. If the maximum number of iterations is reached or the fitness value of the optimal solution in the population changes less than the threshold value in several consecutive generations, the optimal delivery path, node visit sequence and arrival time, as well as the total cost of each path and the hierarchical distribution network cost are output. If the above requirements are not met, return to step d;

[0095] Compared with the prior art, the present invention has the following beneficial technical effects:

[0096] 1. By adopting an optimization strategy that combines the adaptive fireworks algorithm with the quantum genetic algorithm, the parameters of spark generation and quantum bit revolving gate are dynamically adjusted, which significantly improves the algorithm's solution efficiency and global search capabilities in complex optimization problems, thereby more efficiently coping with dynamic environments and complex constraints;

[0097] 2. The insertion algorithm is introduced to optimize the initial solution, which can quickly reduce invalid paths in path planning, optimize the quality of the initial population, effectively shorten the algorithm convergence time, and improve the quality of the overall solution;

[0098] 3. Combined with the adaptive parameter adjustment mechanism, the algorithm can dynamically adjust the number of sparks and the rotation angle of the quantum bit according to the fitness, effectively balancing the global search and local search capabilities of the algorithm to avoid falling into the local optimum;

[0099] 4. Integrate the mutation and crossover operations in quantum genetic algorithms to effectively maintain the diversity of the population, enhance the scalability and robustness of the solution, and be suitable for solving large-scale optimization problems and multi-objective scenarios;

[0100] 5. The algorithm structure has strong modular characteristics, and the weight ratio of the fireworks and quantum genetic parts can be flexibly adjusted according to the specific problem requirements, adapting to the diverse needs of system scalability and application scenarios in future dynamic optimization problems. BRIEF DESCRIPTION OF THE DRAWINGS

[0101] In order to more clearly illustrate the technical solution of the present invention, the specific implementation of the present invention is further described in detail below with reference to the accompanying drawings, wherein:

[0102] Figure 1 is a block diagram of the method of the present invention;

[0103] Figure 2 It is a schematic diagram of the structure of a three-level urban and rural distribution network system based on joint distribution of the present invention;

[0104] Figure 3 It is a schematic diagram of an improved heuristic algorithm flow based on an adaptive fireworks-quantum genetic hybrid algorithm provided by an embodiment of the present invention;

[0105] Figure 4 It is a schematic diagram of a disturbance operator in a fireworks algorithm provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0106] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention are clearly and completely described. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0107] The embodiment of the present application solves the path planning problem of the three-level network for urban and rural joint distribution by providing a method for path planning of the three-level urban and rural joint distribution network based on an adaptive hybrid algorithm.

[0108] The technical solution in the embodiment of the present application is to solve the above technical problems, and the overall idea is as follows:

[0109] In general urban and rural express delivery scenarios, goods are uniformly delivered by various express companies to large urban joint distribution centers. After distribution, they are transported in batches to several township and county transfer stations. Townships and counties have joint distribution transfer stations to temporarily store the express goods in transit, and then uniformly distribute them to the terminal rural self-pickup points. The vehicle path planning problem in this type of distribution scenario can be abstracted as a two-layer vehicle path planning problem. Currently, there is little research on the unified solution of the upper and lower layers. The present invention takes into account the large demand in the urban and rural joint distribution scenario, especially in the upper-layer distribution network, which requires multiple vehicles for distribution, which does not meet the access uniqueness constraint of the general path planning problem. The embodiment of the present invention relaxes this constraint and allows the upper-layer distribution network vehicles to visit satellite sites multiple times, that is, the demand can be split for distribution.

[0110] In addition, since express products have certain delivery time requirements, the embodiment of the present invention is also set as a path planning problem with a soft time window. Therefore, the present invention has not yet received much attention from scholars, but it meets the needs of some delivery scenarios in reality.

[0111] In summary, the present invention can obtain a three-level urban-rural distribution network path planning solution based on joint distribution within an acceptable time, thereby greatly reducing the total cost of urban-rural logistics joint distribution.

[0112] In order to better understand the above technical solution, the above technical solution will be described in detail below in conjunction with the accompanying drawings and specific implementation methods.

[0113] like Figure 1 As shown, the embodiment of the present invention provides a method for path planning of a three-level urban-rural joint distribution network based on an adaptive hybrid algorithm, comprising:

[0114] S1. Obtain vehicle resources, status and delivery task information of the common delivery platform;

[0115] S2. Constructing a mathematical model with the goal of minimizing the total delivery cost based on the vehicle resources and status and the delivery task information;

[0116] S3. An improved heuristic algorithm of an adaptive fireworks-quantum genetic hybrid algorithm is used to solve the mathematical model to obtain the optimal vehicle path of the three-level urban-rural distribution network model, including:

[0117] S31. In the first stage, an initial solution is generated by quantum bit encoding to represent the delivery path. The solution is represented by quantum bits, the position of each delivery node is initialized in the form of a quantum state, multiple potential paths are generated through the superposition state of quantum bits, and the initial solution is selected through quantum measurement.

[0118] S32. In the second stage, in order to improve the quality of the initial solution, an insertion algorithm is used to optimize the initial solution. The insertion algorithm optimizes the initially generated solution by selecting the insertion position with the minimum increment, that is, when inserting an unvisited node into an existing path, selecting the position that minimizes the total cost of the path.

[0119] S33. In the third stage, the fireworks algorithm is used for optimization. The fireworks algorithm works after the initial solution is generated, and uses sparks to explore better solutions in the solution space. Each spark represents a solution, and the number of sparks and the explosion radius are adaptive and gradually adjusted during the iteration process. During the optimization process of the fireworks algorithm, each spark performs a local search for the current solution. By randomly perturbing the spark position, the algorithm can find a better solution in the local solution space.

[0120] S34. In the fourth stage, the quantum genetic algorithm is used to continue to update the optimal solution. In the quantum genetic algorithm, quantum bit encoding is used to represent each individual, that is, each solution. Each individual represents its path through the state of the quantum bit, and the quantum bit encoding can effectively represent the decision space. In each iteration, new solutions are continuously generated through quantum bit encoding and quantum genetic operations, the objective function is optimized, and the path planning is improved.

[0121] S35. In the fifth stage, non-dominated sorting and selection are performed to determine whether the termination condition is met and output the optimal solution.

[0122] In the above scheme, considering the large demand in the urban and rural joint distribution scenario, especially in the upper-level distribution network, multiple vehicles are required for distribution, which does not meet the access uniqueness constraint of the general path planning problem. The embodiment of the present invention relaxes this constraint and allows the upper-level distribution network vehicles to visit satellite sites multiple times, that is, the demand can be split for distribution.

[0123] Jot down the steps that will explain the above scenario in detail:

[0124] In step S1, the vehicle resources and status and delivery task information of the common delivery platform are obtained.

[0125] like Figure 3 As shown, the embodiment of the present invention essentially solves a three-level urban-rural joint distribution network path planning problem based on an adaptive hybrid algorithm. In this problem, all information is determined. The vehicles in the upper distribution network start from the urban joint distribution center, complete the distribution tasks of the township and county transfer stations, and return to the urban joint distribution center. The distribution vehicles in the lower distribution network start from the township and county transfer stations, complete the distribution tasks of the rural terminal self-pickup points, and return to the township and county transfer stations from which they started; each township and county transfer station is allowed to be visited multiple times, and each rural terminal self-pickup point can only be visited once.

[0126] On this basis, and for the needs of subsequent modeling, it can be known that this step requires at least obtaining the following vehicle resources, status, and delivery task information of the same delivery platform:

[0127] D represents the set of urban common distribution centers, S represents the set of township and county transfer stations, and P represents the set of rural terminal pick-up points.

[0128] U represents the set of upper-level distribution vehicles, c1 represents the unit distance transportation cost of upper-level network distribution vehicles, K represents the set of lower-level distribution vehicles, c2 represents the unit distance transportation cost of lower-level network distribution vehicles, v 1is is the speed of the upper-level delivery vehicle u from node i to node s, v 2ij is the speed of the lower-level delivery vehicle k from node i to node j, c3 represents the unit cost coefficient of delay penalty, T i represents the latest arrival time expected by customer i (i.e., the upper limit of the time window).

[0129] z su is the transportation volume when the upper distribution vehicle u visits the transfer station s, Q1 is the maximum capacity constraint of the upper distribution network vehicles, Q2 is the maximum capacity constraint of the lower distribution network vehicles, T iu is the time it takes for the upper-level delivery vehicle u to arrive at node i, d is is the distance between node i and node s, T jk It is the time when the lower-level delivery vehicle k arrives at the rural terminal pick-up point j.

[0130] In step S2, a mathematical model with the goal of minimizing the total delivery cost is constructed based on the vehicle resources and status and the delivery task information;

[0131] The mathematical model includes:

[0132] According to a preferred embodiment of the present invention, according to the basic information, the criterion parameters of vehicle distribution are calculated to establish the objective function of minimum total distribution cost; including:

[0133] The transportation cost of upper-level distribution vehicles, the transportation cost of lower-level distribution vehicles, and the time penalty cost;

[0134] The calculation formula of the upper distribution vehicle transportation cost C1 is:

[0135]

[0136] Among them, c1 represents the unit distance transportation cost of the upper network distribution vehicle, d ij represents the distance from node i to node j, r iju It is a 0-1 variable, indicating whether vehicle u passes through nodes i and j, D represents the set of urban common distribution centers, S represents the set of township and county transfer stations, and U represents the set of upper-level distribution vehicles.

[0137] The calculation formula of the lower-level distribution vehicle transportation cost C2 is:

[0138]

[0139] Among them, c2 represents the unit distance transportation cost of the lower-level network distribution vehicle, d ij represents the distance from node i to node j, x ijk It is a 0-1 variable, indicating whether vehicle k passes through nodes i and j, S represents the set of township and county transfer stations, P represents the set of rural terminal pick-up points, and K represents the set of lower-level distribution vehicles.

[0140] The calculation formula of the time penalty cost c3 is:

[0141]

[0142] Where c3 represents the unit cost coefficient of delay penalty, T i represents the latest arrival time expected by customer i (i.e., the upper limit of the time window), t i represents the time point when the vehicle arrives at the rural terminal self-pickup point i, and P represents the set of rural terminal self-pickup points;

[0143] According to the above criteria parameters, the objective function of minimizing the total distribution cost is established, which is expressed as:

[0144]

[0145] Taking the delivery time of the terminal node, the coherence of the delivery path, the vehicle capacity, and the node demand satisfaction as the constraints of the objective function of the minimum total delivery cost, a three-level urban-rural distribution network model based on joint distribution is constructed, including:

[0146]

[0147] In the formula, z su is the transportation volume when the upper-level distribution vehicle u visits the transfer station s, Q1 is the maximum capacity constraint of the upper-level distribution network vehicles;

[0148]

[0149] In the formula, q i is the demand of the terminal rural pick-up point i, Q2 is the maximum capacity constraint of the lower-level distribution network vehicles;

[0150]

[0151] In the formula,

[0152]

[0153] In the formula,

[0154]

[0155] In the formula, e s is the demand for transfer stations in townships and counties;

[0156]

[0157]

[0158] Where M is a sufficiently large real number;

[0159] T iu +d is / v 1is +M×(1-r isu )≥T su i∈D∪S,s∈S,u∈U (15)

[0160] T iu +d is / v 1is -M×(1-r isu )≤T su i∈D∪S,s∈S,u∈U (16)

[0161] Where, T iu is the time it takes for the upper-level delivery vehicle u to arrive at node i, d is is the distance between node i and node s, v 1is is the speed of the upper-level delivery vehicle u from node i to node s, T su is the time when the upper-level delivery vehicle u arrives at the township or county transfer station s;

[0162] T ik +d ij / v 2ij +max(T su )+M×(1-x ijk )≥T jk i∈S∪P,s∈S,j∈P,k∈K (17)

[0163] T ik +d ij / v 2ij +max(T su )-M×(1-x ijk )≤T jk j∈S∪P,s∈S,j∈P,k∈K (18)

[0164] Where, Tik is the time it takes for the lower-level delivery vehicle k to arrive at node i, d ij is the distance between node i and node j, v 2ij is the speed of the lower-level delivery vehicle k from node i to node j, T jk is the time it takes for the lower-level delivery vehicle k to arrive at the rural terminal pick-up point j;

[0165] t i =max(t ik ) (19)

[0166] Where, t i The time it takes for the lower-level delivery vehicle k to arrive at the rural terminal pickup point i;

[0167] in,

[0168] Formula (5) represents the maximum load constraint of vehicles in the upper distribution network;

[0169] Formula (6) represents the maximum load constraint of the vehicles in the lower-level distribution network, and Q2 is the maximum capacity constraint of the vehicles in the lower-level distribution network;

[0170] Equations (7) and (8) indicate that in the upper distribution network, each node j must have only one vehicle reaching it from other nodes, and each node i must have only one vehicle departing from it, ensuring the integrity and continuity of the path;

[0171] Formula (9) and Formula (10) indicate that each terminal rural pick-up point must be visited by a vehicle and can only be visited and served once and only one vehicle can depart from it;

[0172] Formula (11) indicates that the demand distribution of the township and county transfer stations is completed;

[0173] Formula (12) indicates that each vehicle in the upper distribution network starts from the urban common distribution center, passes through several township and county transfer stations, and then returns to the urban common distribution center;

[0174] Formula (13) indicates that each vehicle in the lower-level distribution network starts from a township or county transfer station, passes through several terminal rural pick-up points, and then returns to the township or county transfer station where it started;

[0175] Formula (14) represents the cargo demand of the township and county transfer stations;

[0176] Formula (15)-Formula (19) represents the time when the goods arrive at the demand point;

[0177] In step S3, an improved heuristic algorithm of an adaptive fireworks-quantum genetic hybrid algorithm is used to solve the optimal path optimization model to determine the optimal vehicle delivery path.

[0178] This embodiment proposes a three-level urban-rural distribution network path planning framework based on the joint distribution model to solve the vehicle path planning problem, which can obtain a three-level urban-rural distribution network path planning solution based on joint distribution within an acceptable time, and greatly reduce the total cost of urban-rural logistics joint distribution, including:

[0179] like Figure 3 As shown, in each stage:

[0180] S31. Perform quantum bit encoding on each township and county transfer station and terminal rural node. Each node or terminal demand point is represented by one quantum bit:

[0181] |ψ i >=α i |0>+β i |1> (20)

[0182] Among them, |α i | 2 and |β i | 2 They represent the probability of node selection state, respectively. |0> means the node is not selected, and |1> means the node is selected. A path is represented by a combination of multiple quantum bits.

[0183] S32, initialize the population and randomly set the initial amplitude α of each quantum bit i and β i , guarantee |α i | 2 +|β i | 2 =1, and then the quantum state |ψ i >Decode to a specific selection value of 0 or 1, combine the selected nodes into a path, assign the unselected nodes to other vehicles, randomly generate N candidate initial solutions based on the superposition state of the quantum bit, check the legitimacy of each path after decoding, and repair the solutions that do not meet the constraints;

[0184] And the parameter settings of the following algorithms can also be given as examples:

[0185] Define the size of the population. For example, set the population size to N = 50, which means that 50 candidate solutions are generated in the initial stage. Each individual is represented by a quantum bit string. The initial state can be uniformly distributed and set to In the parameter initialization of the fireworks algorithm, the maximum number of sparks is M max =20, in the initialization of the quantum genetic algorithm parameters, the initial value of the angle step of the revolving door is set to Δθ = 0.05π, and the disturbance range of the spark generation is set to The termination condition of the entire algorithm sets the maximum number of iterations T max=200, or the change in the objective function is less than the threshold ∈=10 -6 .

[0186] S33, optimize the initial solution, use the insertion algorithm to optimize the initial solution, reduce the path cost and improve the feasibility of the solution, for each vehicle path in the initial solution, traverse the unvisited nodes in turn, insert the unvisited node i into all possible positions in the path, and calculate the path cost increment of each insertion scheme:

[0187] ΔC=C new -C original (twenty one)

[0188] Among them, C new represents the path cost after inserting the node, C original Represents the path cost before inserting the node. In order to increase the diversity of the initial population, the present invention selects the position with the smallest cost increment to insert node i and updates the path. After all nodes are inserted, the optimized path is obtained. Check whether the optimized path meets the constraints of capacity and time window. If not, adjust the node order or reallocate the node to other vehicle paths to ensure that the optimized initial solution is feasible.

[0189] S34, using the fireworks algorithm to perform dynamic global search, first calculate the number of sparks N at the center of each firework explosion using formula (22) s , where N max ,N min Respectively represent the maximum and minimum number of sparks, F(x) represents the fitness value of the current solution x, reflecting the quality of the solution. Then, for each individual vehicle, an explosion operation is used to generate sparks to determine the spark position. The spark generation position is determined by the disturbance range R, which is determined by formula (23), where R represents the explosion radius, and Δx represents the random disturbance factor, which is randomly generated in the range [-1,1]. Calculate each spark solution x s The fitness value F(x s ), for the fireworks center with better fitness, a smaller explosion radius is set to generate a fine spark solution; for the fireworks center with poor fitness, a larger explosion radius is set to expand the search range;

[0190]

[0191] x s =x+R·Δx,Δx∈[-1,1] (23)

[0192] like Figure 4 As shown, the present invention perturbs the node or path of the current path solution x to generate N s Spark solutions, design the following three perturbation operators:

[0193] Node exchange operator: randomly obtain two nodes in the path, exchange their positions, and generate a new path solution;

[0194] Node insertion operator: Using equation (24), move a node i in the path to another position j;

[0195] x′=Insert(x,i,j) (24)

[0196] Path splitting operator: splits a path into two parts and redistributes them to different vehicles.

[0197] Preferably, if the spark solution does not satisfy the constraints, its fitness value is reduced by a penalty function:

[0198] F penalized =F(x s )+λ·violation_penalty

[0199] Preferably, among the newly generated spark solutions, the spark solution with the highest fitness is selected as the new firework center, and the fitness F(x new ) and the global optimal solution F best , if the current solution is better, update the global optimal solution:

[0200] F best =min(F best ,F(x new )) (25)

[0201] Preferably, retain N optimal firework centers as the initial solution for the next round of iteration;

[0202] S35. Use the quantum genetic algorithm to locally optimize the solution of the fireworks algorithm. First, use equation (26) to dynamically adjust the rotation angle θ according to the fitness to perform quantum revolving door to update the population:

[0203]

[0204] Preferably, the rotation amplitude is dynamically adjusted according to the fitness value of the solution, so as to locally strengthen the high-quality solution and increase the randomness of the low-quality solution, so as to avoid falling into the local optimum;

[0205]

[0206] Preferably, the following two operators are designed:

[0207] Crossover operator: randomly select two path solutions from the population, exchange some of their nodes, and generate a new path solution that inherits the excellent characteristics of the parent generation:

[0208] x new=Crossover(x1,x2) (29)

[0209] Mutation operator: randomly adjusts the order of nodes in the path, randomly selects two nodes in the path to swap positions, and randomly inserts or deletes a node;

[0210] Preferably, the fitness values ​​of the solutions after crossover and mutation are calculated, solutions with higher fitness values ​​are retained, solutions with lower fitness values ​​are eliminated, the newly generated solutions are merged with the current population, the N solutions with the highest fitness values ​​are selected as the new population and the global optimal solution is updated;

[0211] Preferably, if the maximum number of iterations is reached or the fitness value of the optimal solution in the population changes less than a threshold value in several consecutive generations, the optimal delivery path, node visit sequence and arrival time, and the total cost corresponding to each path and the hierarchical delivery network cost are output. If the above requirements are not met, repeat step de;

[0212] Compared with the prior art, the present invention has the following beneficial technical effects:

[0213] 1. By adopting an optimization strategy that combines the adaptive fireworks algorithm with the quantum genetic algorithm, the parameters of spark generation and quantum bit revolving gate are dynamically adjusted, which significantly improves the algorithm's solution efficiency and global search capabilities in complex optimization problems, thereby more efficiently coping with dynamic environments and complex constraints;

[0214] 2. The insertion algorithm is introduced to optimize the initial solution, which can quickly reduce invalid paths in path planning, optimize the quality of the initial population, effectively shorten the algorithm convergence time, and improve the quality of the overall solution;

[0215] 3. Combined with the adaptive parameter adjustment mechanism, the algorithm can dynamically adjust the number of sparks and the rotation angle of the quantum bit according to the fitness, effectively balancing the global search and local search capabilities of the algorithm to avoid falling into the local optimum;

[0216] 4. Integrate the mutation and crossover operations in quantum genetic algorithms to effectively maintain the diversity of the population, enhance the scalability and robustness of the solution, and be suitable for solving large-scale optimization problems and multi-objective scenarios;

[0217] 5. The algorithm structure has strong modular characteristics, and the weight ratio of the fireworks and quantum genetic parts can be flexibly adjusted according to the specific problem requirements, adapting to the diverse needs of system scalability and application scenarios in future dynamic optimization problems.

[0218] The above description is only a preferred embodiment of the present invention and does not limit the present invention in any way. Any simple modification, change and equivalent transformation made to the above embodiment based on the technical essence of the present invention still falls within the protection scope of the technical solution of the present invention.

Claims

1. A method for path planning of a three-level urban-rural joint distribution network based on an adaptive hybrid algorithm, which is characterized by: The steps include: S1. Obtain vehicle resources, status and delivery task information of the common delivery platform; S2. constructing a mathematical model with the goal of minimizing the total delivery cost based on the vehicle resources and status and delivery task information; S3. An improved heuristic algorithm of an adaptive fireworks-quantum genetic hybrid algorithm is used to solve the mathematical model and obtain the optimal vehicle path of the three-level urban-rural distribution network model, including: In the first stage, an initial solution is generated by quantum bit encoding to represent the delivery path. Using quantum bits to represent the solution, the position of each delivery node is initialized in the form of a quantum state, multiple potential paths are generated through the superposition state of quantum bits, and the initial solution is selected through quantum measurement. In the second stage, in order to improve the quality of the initial solution, the insertion algorithm is used to optimize the initial solution. The insertion algorithm optimizes the initially generated solution by selecting the insertion position with the minimum increment, that is, when inserting an unvisited node into an existing path, the position that minimizes the total cost of the path is selected. In the third stage, the fireworks algorithm is used for optimization. The fireworks algorithm works after the initial solution is generated, and uses sparks to explore better solutions in the solution space. Each spark represents a solution, and the number of sparks and the explosion radius are adaptive and gradually adjusted during the iteration process. During the optimization process of the fireworks algorithm, each spark performs a local search for the current solution. By randomly perturbing the spark position, the algorithm can find a better solution in the local solution space. In the fourth stage, the quantum genetic algorithm is used to continue to update the optimal solution. In the quantum genetic algorithm, quantum bit encoding is used to represent each individual, that is, each solution. Each individual represents its path through the state of the quantum bit, and the quantum bit encoding can effectively represent the decision space. In each iteration, new solutions are continuously generated through quantum bit encoding and quantum genetic operations, optimizing the objective function and improving path planning. In the fifth stage, non-dominated sorting and selection, it is determined whether the termination conditions are met and the optimal solution is output.

2. The method for path planning of a three-level urban-rural joint distribution network as claimed in claim 1, characterized in that: The mathematical model in step S2 includes: The transportation cost of upper-level distribution vehicles, the transportation cost of lower-level distribution vehicles, and the time penalty cost; The calculation formula of the upper distribution vehicle transportation cost C1 is: Among them, c1 represents the unit distance transportation cost of the upper network distribution vehicle, d ij represents the distance from node i to node j, r iju It is a 0-1 variable, indicating whether vehicle u passes through nodes i and j, D represents the set of urban common distribution centers, S represents the set of township and county transfer stations, and U represents the set of upper-level distribution vehicles. The calculation formula of the lower-level distribution vehicle transportation cost C2 is: Among them, c2 represents the unit distance transportation cost of the lower-level network distribution vehicle, d ij represents the distance from node i to node j, x ijk It is a 0-1 variable, indicating whether vehicle k passes through nodes i and j, S represents the set of township and county transfer stations, P represents the set of rural terminal pick-up points, and K represents the set of lower-level distribution vehicles. The calculation formula of the time penalty cost c3 is: Where c3 represents the unit cost coefficient of delay penalty, T i represents the latest arrival time expected by customer i (i.e., the upper limit of the time window), t i represents the time point when the vehicle arrives at the rural terminal self-pickup point i, and P represents the set of rural terminal self-pickup points; According to the above criteria parameters, the objective function of minimizing the total distribution cost is established, which is expressed as:

3. The method for path planning of a three-level urban-rural joint distribution network as claimed in claim 2, characterized in that: The mathematical model also includes: Constraints: Taking the delivery time of the terminal node, the coherence of the delivery path, the vehicle capacity, and the node demand satisfaction as the constraints of the objective function of the minimum total delivery cost, a three-level urban-rural distribution network model based on joint distribution is constructed, including: In the formula, z su is the transportation volume when the upper-level distribution vehicle u visits the transfer station s, Q1 is the maximum capacity constraint of the upper-level distribution network vehicles; In the formula, q i is the demand of the terminal rural pick-up point i; In the formula, In the formula, In the formula, e s is the demand for transfer stations in townships and counties; Where M is a sufficiently large real number; T iu +d is / v 1is +M×(1-r isu )≥T su i∈D∪S,s∈S,u∈U (15) T iu +d is / v 1is -M×(1-r isu )≤T su i∈D∪S,s∈S,u∈U (16) Where, T iu is the time it takes for the upper-level delivery vehicle u to arrive at node i, d is is the distance between node i and node s, v 1is is the speed of the upper-level delivery vehicle u from node i to node s, T su is the time when the upper-level delivery vehicle u arrives at the township or county transfer station s; T ik +d ij / v 2ij +max(T su )+M×(1-x ijk )≥T jk i∈S∪P,s∈S,j∈P,k∈K (17) T ik +d ij / v 2ij +max(T su )-M×(1-x ijk )≤T jk i∈S∪P,s∈S,j∈P,k∈K (18) Where, T ik is the time it takes for the lower-level delivery vehicle k to arrive at node i, d ij is the distance between node i and node j, v 2ij is the speed of the lower-level delivery vehicle k from node i to node j, T jk is the time it takes for the lower-level delivery vehicle k to arrive at the rural terminal pick-up point j; t i =max(t ik ) (19) Where, t i The time it takes for the lower-level delivery vehicle k to arrive at the rural terminal pickup point i; in, Formula (5) represents the maximum load constraint of vehicles in the upper distribution network; Formula (6) represents the maximum load constraint of the vehicles in the lower-level distribution network, and Q2 is the maximum capacity constraint of the vehicles in the lower-level distribution network; Equations (7) and (8) indicate that in the upper distribution network, each node j must have only one vehicle reaching it from other nodes, and each node i must have only one vehicle departing from it, ensuring the integrity and continuity of the path; Formula (9) and Formula (10) indicate that each terminal rural pick-up point must be visited by a vehicle and can only be visited and served once and only one vehicle can depart from it; Formula (11) indicates that the demand distribution of the township and county transfer stations is completed; Formula (12) indicates that each vehicle in the upper distribution network starts from the urban common distribution center, passes through several township and county transfer stations, and then returns to the urban common distribution center; Formula (13) indicates that each vehicle in the lower-level distribution network starts from a township or county transfer station, passes through several terminal rural pick-up points, and then returns to the township or county transfer station where it started; Formula (14) represents the cargo demand of the township and county transfer stations; Formula (15)-Formula (19) represents the time when the goods arrive at the demand point.

4. The method for path planning of a three-level urban-rural joint distribution network as described in claim 3, characterized in that: Step S3 includes: S31. Perform quantum bit encoding on each township and county transfer station and terminal rural node. Each node or terminal demand point is represented by one quantum bit: |ψ i >=a i |0>+β i |1> (20) Among them, |α i | 2 and |β i | 2 They represent the probability of node selection state, respectively. |0> means the node is not selected, and |1> means the node is selected. A path is represented by a combination of multiple quantum bits. S32, initialize the population and randomly set the initial amplitude α of each quantum bit i and β i , guarantee |α i | 2 +|β i | 2 =1, and then the quantum state |ψ i >Decode to a specific selection value of 0 or 1, combine the selected nodes into a path, assign the unselected nodes to other vehicles, randomly generate N candidate initial solutions based on the superposition state of the quantum bit, check the legitimacy of each path after decoding, and repair the solutions that do not meet the constraints; S33, optimize the initial solution, use the insertion algorithm to optimize the initial solution, reduce the path cost and improve the feasibility of the solution, for each vehicle path in the initial solution, traverse the unvisited nodes in turn, insert the unvisited node i into all possible positions in the path, and calculate the path cost increment of each insertion scheme: ΔC=C new -C original (21) Among them, C new represents the path cost after inserting the node, C original Represents the path cost before inserting the node. In order to increase the diversity of the initial population, the present invention selects the position with the smallest cost increment to insert node i and updates the path. After all nodes are inserted, the optimized path is obtained. Check whether the optimized path meets the constraints of capacity and time window. If not, adjust the node order or reallocate the node to other vehicle paths to ensure that the optimized initial solution is feasible. After the initial solution is optimized, calculate the fitness value of each path and update the optimal solution of the group. S34, using the fireworks algorithm to perform dynamic global search, first calculate the number of sparks N at the center of each firework explosion using formula (22) s , where N max , N min Respectively represent the maximum and minimum number of sparks, F(x) represents the fitness value of the current solution x, reflecting the quality of the solution. Then, for each individual vehicle, an explosion operation is used to generate sparks to determine the spark position. The spark generation position is determined by the disturbance range R, which is determined by formula (23), where R represents the explosion radius, and Δx represents the random disturbance factor, which is randomly generated in the range [-1,1]. Calculate each spark solution x s The fitness value F(x s ), for the fireworks center with better fitness, a smaller explosion radius is set to generate a fine spark solution; for the fireworks center with poor fitness, a larger explosion radius is set to expand the search range; x s =x+R·Δx,Δx∈[-1,1] (23) Perturb the nodes or paths of the current path solution x to generate N s Spark solutions, design the following three perturbation operators: Node exchange operator: randomly obtain two nodes in the path, exchange their positions, and generate a new path solution; Node insertion operator: Using equation (24), move a node i in the path to another position j; x′=Insert(x,i,j) (24) Path splitting operator: splits a path into two parts and redistributes them to different vehicles. Preferably, if the spark solution does not satisfy the constraints, its fitness value is reduced by a penalty function: F penalized =F(x s )+λ·violation_penalty Preferably, among the newly generated spark solutions, the spark solution with the highest fitness is selected as the new firework center, and the fitness F(x new ) and the global optimal solution F best , if the current solution is better, update the global optimal solution: F best =min(F best ,F(x new )) (25) Preferably, retain N optimal firework centers as the initial solution for the next round of iteration; S35. Use the quantum genetic algorithm to locally optimize the solution of the fireworks algorithm. First, use equation (26) to dynamically adjust the rotation angle θ according to the fitness to perform quantum revolving door to update the population: Preferably, the rotation amplitude is dynamically adjusted according to the fitness value of the solution, so as to locally strengthen the high-quality solution and increase the randomness of the low-quality solution, so as to avoid falling into the local optimum; Preferably, the following two operators are designed: Crossover operator: randomly select two path solutions from the population, exchange some of their nodes, and generate a new path solution that inherits the excellent characteristics of the parent generation: x new =Crossover(x1,x2) (29) Mutation operator: randomly adjusts the order of nodes in the path, randomly selects two nodes in the path to swap positions, and randomly inserts or deletes a node; Calculate the fitness value of the solution after crossover and mutation, retain the solution with higher fitness value, eliminate the solution with lower fitness value, merge the newly generated solution with the current population, select the N solutions with the highest fitness as the new population and update the global optimal solution; S36. If the maximum number of iterations is reached or the fitness value of the optimal solution in the population changes by less than a threshold value in several consecutive generations, the optimal delivery path, node visit sequence and arrival time, as well as the total cost corresponding to each path and the hierarchical distribution network cost are output; if the above requirements are not met, return to step S34.

Images