Cold-chain logistics path planning method and device, electronic equipment and storage medium

Abstract

The invention provides a cold-chain logistics path planning method and device, electronic equipment and a storage medium, and belongs to the field of logistics distribution. The method comprises the steps of obtaining distribution related information of to-be-planned cold-chain logistics and to-be-planned variables, and constructing an objective function and constraint conditions; initializing a variable to be planned to obtain an initial particle swarm; determining particle fitness according to a target function; constructing a candidate solution equalization pool based on a reverse learning strategy; generating a transition vector based on a random selection updating strategy, and performing Gaussian refresh strategy population updating on the initial particle swarm according to the transition vector to obtain a new round of particle swarm; and iterating the updating process to obtain an optimal particle position, and determining an optimal cold-chain logistics path. According to the method, the population diversity is effectively expanded through a reverse learning strategy, the search space is increased, the algorithm search capability is improved through a random selection update strategy, particles jump out of local optimum through a Gaussian refresh strategy, the capability of comprehensively exploring the optimal solution of the algorithm can be effectively improved, and falling into local optimum is avoided.

Description

Technical Field

[0001] The present invention relates to the field of logistics and distribution technology, and in particular to a cold chain logistics path planning method, device, electronic equipment and storage medium. Background Art

[0002] In the field of cold chain logistics, reasonable route planning can effectively improve logistics efficiency, reduce costs and ensure the quality of goods. However, due to the particularity of transported goods, cold chain logistics must not only consider transportation costs and time efficiency, but also strictly control temperature and meet customer time window requirements. Traditional optimization algorithms often find it difficult to find the optimal solution when dealing with the multi-objective and multi-constraint problem of cold chain logistics route planning.

[0003] Swarm intelligence optimization algorithms can optimize and solve some complex problems by simulating natural populations, providing a new approach for cold chain logistics path planning. However, when existing swarm intelligence algorithms are applied to cold chain logistics path planning, due to the large number of factors that need to be considered in cold chain logistics path planning, the existing swarm intelligence optimization algorithms tend to aggregate population particles too early during the iterative optimization process, resulting in a limited search space. It is easy to fall into local optimality in cold chain logistics path planning, making it difficult to fully explore the optimal solution.

[0004] Therefore, when the existing swarm intelligence algorithm is used for cold chain logistics path planning, there is a technical problem that the swarm particles tend to focus too early, resulting in a limited search space, which in turn causes the solution process to fall into a local optimum and makes it difficult to fully explore the optimal solution. This needs to be improved. Summary of the invention

[0005] In view of this, it is necessary to provide a cold chain logistics path planning method, device, electronic device and storage medium to solve the technical problem in the prior art that population particles are prone to premature focusing, resulting in limited search space, and then causing the solution process to fall into local optimality, making it difficult to fully explore the optimal solution.

[0006] In order to solve the above problems, on the one hand, the present invention provides a cold chain logistics path planning method, comprising: Obtain the distribution-related information and variables to be planned for the cold chain logistics to be planned, and construct the objective function and constraint conditions for the cold chain logistics path planning based on the distribution-related information and variables to be planned; Initialize the variables to be planned to obtain the initial particle swarm, determine the particle fitness according to the objective function, use the reverse learning strategy to build the candidate solution equilibrium pool according to the particle fitness, generate the transition vector by randomly selecting the update strategy according to the candidate solution equilibrium pool, and update the initial particle swarm with the Gaussian refresh strategy according to the transition vector to obtain a new round of particle swarm. The optimal particle position is obtained through the iterative update process, and the optimal cold chain logistics path is determined according to the optimal particle position.

[0007] In one possible implementation, the delivery-related information includes at least one of target location information, delivery demand, delivery time window requirement, initial delivery location and vehicle information, and the variables to be planned include at least one of delivery vehicle, delivery route and delivery start time.

[0008] In one possible implementation, the objective function includes minimizing at least one of vehicle driving cost, cooling cost, carbon emission cost and penalty cost, and the constraint condition includes at least one of vehicle load constraint and time window constraint.

[0009] In a possible implementation, initializing the variables to be planned to obtain an initial particle swarm includes: Determine the particle search space boundary according to the constraint conditions; Initial particles are constructed within the boundary of the particle search space based on preset adaptive weights to obtain an initial particle swarm.

[0010] In a possible implementation, a candidate solution balancing pool is constructed by performing a reverse learning strategy based on particle fitness, including: Generate a mirror particle swarm according to the particle positions of the initial particle swarm and the boundary of the particle search space; After merging the initial particle swarm and the mirror particle swarm, fitness screening is performed to obtain several elite solutions, and a candidate solution balance pool is constructed based on the elite solution and the average value of the elite solution.

[0011] In a possible implementation, a transition vector is generated by performing a random selection update strategy based on a candidate solution balancing pool, including: The first candidate solution is obtained by random selection in the candidate solution balance pool, and the second candidate solution is obtained by random selection in the initial particle swarm; A first candidate solution or a second candidate solution is randomly selected based on a preset selection probability, and based on a preset first position update strategy corresponding to the first candidate solution or a preset second position update strategy corresponding to the second candidate solution, a particle position optimization is performed on the particles to be updated of the initial particle swarm to obtain a transition vector.

[0012] In a possible implementation, the initial particle swarm is updated with a Gaussian refresh strategy population according to the transition vector to obtain a new round of particle swarm, including: Determine the updated fitness of the transition vector and the initial fitness of the particle to be updated according to the objective function; If the updated fitness is greater than the initial fitness, the transition vector is used as a new round of particles in the particle swarm; If the updated fitness is less than or equal to the initial fitness, the particle to be updated is updated by Gaussian mutation based on the preset mutation probability and then used as the new round of particles in the particle swarm.

[0013] On the other hand, the present invention also provides a cold chain logistics path planning device, comprising: A planning condition construction unit is used to obtain the distribution-related information and variables to be planned for the cold chain logistics to be planned, and to construct the objective function and constraint conditions for the cold chain logistics path planning according to the distribution-related information and variables to be planned; The iterative optimization solving unit is used to initialize the variables to be planned to obtain the initial particle swarm, determine the particle fitness according to the objective function, build a candidate solution equilibrium pool based on the reverse learning strategy according to the particle fitness, generate a transition vector based on the random selection update strategy of the candidate solution equilibrium pool, and update the initial particle swarm with the Gaussian refresh strategy according to the transition vector to obtain a new round of particle swarm. The iterative update process obtains the optimal particle position, and determines the optimal cold chain logistics path based on the optimal particle position.

[0014] On the other hand, the present invention also provides an electronic device, including a processor, a memory, and a computer program stored in the memory and executable on the processor. When the processor executes the program, the above-mentioned cold chain logistics path planning method is implemented.

[0015] On the other hand, the present invention also provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the above-mentioned cold chain logistics path planning method is implemented.

[0016] The beneficial effects of the present invention are as follows: in the cold chain logistics path planning method provided by the present invention, firstly, the distribution-related information and variables to be planned of the cold chain logistics to be planned are obtained, and the objective function and constraint conditions of the cold chain logistics path planning are constructed according to the distribution-related information and the variables to be planned; then, the variables to be planned are initialized to obtain the initial particle group, the particle fitness is determined according to the objective function, the candidate solution balance pool is constructed according to the particle fitness by the reverse learning strategy, the transition vector is generated by the random selection update strategy according to the candidate solution balance pool, the initial particle group is updated by the Gaussian refresh strategy population according to the transition vector to obtain a new round of particle group, the optimal particle position is obtained by the iterative update process, and the optimal cold chain logistics path is determined according to the optimal particle position. The present invention effectively expands the population diversity and increases the search space through the reverse learning strategy, improves the algorithm search ability through the random selection update strategy, and makes the particles jump out of the local optimum through the Gaussian refresh strategy, which can effectively improve the algorithm's ability to fully explore the optimal solution and avoid falling into the local optimum. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative work.

[0018] Figure 1 A schematic diagram of a process flow of an embodiment of a cold chain logistics path planning method provided by the present invention; Figure 2 A schematic diagram of a process for constructing an initial particle swarm according to an embodiment of the present invention; Figure 3 A schematic diagram of a process for constructing a candidate solution balancing pool by a reverse learning strategy according to an embodiment of the present invention; Figure 4 A schematic diagram of a flow chart of a random selection update strategy according to an embodiment of the present invention; Figure 5 A schematic diagram of a process flow of a Gaussian refresh strategy provided by an embodiment of the present invention; Figure 6 It is a fitness iteration curve diagram of an embodiment of the present invention; Figure 7 A schematic diagram of the structure of an embodiment of a cold chain logistics path planning device provided by the present invention; Figure 8 A schematic structural diagram of an embodiment of an electronic device provided by the present invention. DETAILED DESCRIPTION

[0019] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only 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 those skilled in the art without creative work are within the scope of protection of the present invention.

[0020] In the description of the embodiments of the present invention, unless otherwise specified, "multiple" means two or more than two. "And / or" describes the association relationship of associated objects, indicating that three relationships may exist. For example, "A and / or B" may mean: A exists alone, A and B exist at the same time, and B exists alone.

[0021] The descriptions of "first" and "second" in the embodiments of the present invention are only used for descriptive purposes and cannot be understood as indicating or implying their relative importance or implicitly indicating the number of the indicated technical features. Therefore, the technical features defined as "first" and "second" may explicitly or implicitly include at least one of the features.

[0022] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present invention. The appearance of the phrase in various places in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0023] The present invention provides a cold chain logistics path planning method, device, electronic device and storage medium, which are described below respectively.

[0024] Figure 1 A schematic diagram of a process flow of an embodiment of the cold chain logistics path planning method provided by the present invention is as follows Figure 1 As shown, the cold chain logistics path planning method includes: S101, obtaining the distribution-related information and variables to be planned for the cold chain logistics, and constructing the objective function and constraint conditions for the cold chain logistics path planning according to the distribution-related information and the variables to be planned; S102, initialize the variables to be planned to obtain the initial particle swarm, determine the particle fitness according to the objective function, use the reverse learning strategy to build a candidate solution equilibrium pool according to the particle fitness, generate a transition vector by randomly selecting and updating the candidate solution equilibrium pool, and update the initial particle swarm with the Gaussian refresh strategy according to the transition vector to obtain a new round of particle swarm, obtain the optimal particle position through the iterative update process, and determine the optimal cold chain logistics path according to the optimal particle position.

[0025] Compared with the prior art, the cold chain logistics path planning method provided by the embodiment of the present invention first obtains the distribution-related information and variables to be planned of the cold chain logistics to be planned, and constructs the objective function and constraints of the cold chain logistics path planning according to the distribution-related information and the variables to be planned; then initializes the variables to be planned to obtain the initial particle group, determines the particle fitness according to the objective function, performs a reverse learning strategy according to the particle fitness to construct a candidate solution balance pool, performs a random selection update strategy according to the candidate solution balance pool to generate a transition vector, performs a Gaussian refresh strategy population update on the initial particle group according to the transition vector to obtain a new round of particle group, obtains the optimal particle position during the iterative update process, and determines the optimal cold chain logistics path according to the optimal particle position. The present invention effectively expands the population diversity and increases the search space through the reverse learning strategy, improves the algorithm search capability through the random selection update strategy, and enables particles to jump out of the local optimum through the Gaussian refresh strategy, which can effectively improve the algorithm's ability to fully explore the optimal solution and avoid falling into the local optimum.

[0026] In some embodiments of the present invention, the delivery-related information includes at least one of target location information, delivery demand, delivery time window requirement, initial delivery location and vehicle information, and the variables to be planned include at least one of delivery vehicle, delivery route and delivery start time.

[0027] Specifically, the embodiment obtains the target location information of the customer receiving goods in the cold chain logistics, the delivery demand for fresh products, the delivery time window requirements for the customer to receive goods, and the initial delivery location and vehicle information of the distribution center by investigating the cold chain logistics distribution area. The delivery vehicle corresponding to each product, the corresponding delivery path and delivery time are used as the variables to be planned in the path planning algorithm.

[0028] In some embodiments of the present invention, the objective function includes minimizing at least one of vehicle driving cost, cooling cost, carbon emission cost and penalty cost, and the constraints include at least one of vehicle load constraint and time window constraint.

[0029] Specifically, the objective function of cold chain logistics path planning in the embodiment is to minimize vehicle driving cost, refrigeration cost, carbon emission cost and penalty cost. Among them, the vehicle driving cost is calculated based on the length of the distribution path and the unit distance cost, and the formula is expressed as ; For cooling costs, based on the vehicle’s stay time at each node and refrigeration unit cost Calculation, the formula is expressed as ; For carbon emission costs, according to carbon emissions And the carbon trading price is obtained, the formula is expressed as ; For penalty cost, it is based on the time it takes for the vehicle to arrive at the customer point With the latest time If the relationship between > , then the penalty cost ,otherwise The total cost is the sum of all costs, and the formula is: .

[0030] It should be noted that the above variables to be planned, objective functions and constraints are only one way of applying the present invention in a certain cold chain logistics path planning embodiment. In other embodiments, such as other cold chain logistics path planning with special requirements, or other similar complex planning problems, fewer or other variables to be planned, objective functions and constraints may be set as needed.

[0031] In some embodiments of the present invention, Figure 2 FIG. 1 is a flow chart of constructing an initial particle swarm according to an embodiment of the present invention. Figure 2As shown, the variables to be planned are initialized to obtain the initial particle swarm, including: S201, determining the particle search space boundary according to the constraint conditions; S202: constructing initial particles within the boundary of the particle search space based on preset adaptive weights to obtain an initial particle swarm.

[0032] Specifically, the embodiment of the present invention is based on the balanced optimizer algorithm, and is combined with the existing swarm intelligence algorithms faced by the cold chain logistics path planning problem. The problem is that the population particles are prone to premature focusing, resulting in a limited search space, which in turn makes the solution process fall into a local optimum and makes it difficult to fully explore the optimal solution problem. An improved balanced optimizer algorithm is obtained.

[0033] In order to improve the algorithm's global search capability, the embodiment introduces an adaptive weight strategy. At the beginning of the iteration, the adaptive weight strategy assigns a larger weight value to the algorithm, so that the particles can make large strides in the search space and widely explore possible optimal areas. It can be expressed as:

[0034] in, is the current iteration number, is the maximum number of iterations. As the iteration proceeds, the weight value decreases nonlinearly according to this formula, guiding the particles to gradually focus on the known better area for fine search, thereby improving the convergence speed and accuracy.

[0035] In the initial stage, the embodiment determines the boundary of the particle search space according to the constraint conditions, and randomly selects initial particles in the search space with a larger weight to obtain an initial particle group.

[0036] In some embodiments of the present invention, Figure 3 The flowchart of the reverse learning strategy of the embodiment of the present invention to construct a candidate solution balancing pool is as follows: Figure 3 As shown, a reverse learning strategy is used according to particle fitness to build a candidate solution equilibrium pool, including: S301, generating a mirror particle swarm according to the particle positions of the initial particle swarm and the particle search space boundary; S302, after merging the initial particle swarm and the mirror particle swarm, fitness screening is performed to obtain several elite solutions, and a candidate solution balancing pool is constructed according to the elite solutions and the average values ​​of the elite solutions.

[0037] Specifically, in order to expand population diversity and increase the search space during the algorithm optimization process, the embodiment constructs a candidate solution balancing pool of the balanced optimizer algorithm through a reverse learning strategy.

[0038] Among them, for the particle population, according to the particle position , the embodiment uses the following formula to find the mirror particle position of each particle in the search space :

[0039] in, and The search space The upper and lower boundaries of the dimension.

[0040] After merging the initial particle group and the mirror particle group, the embodiment calculates the fitness value of each particle, sorts and screens them according to the fitness value, takes several particles with higher rankings as elite solutions, and constructs a candidate solution balance pool through these particles and their average values. Through this operation, the embodiment fully utilizes the information of the elite solution, effectively expands the population diversity, effectively avoids the algorithm from falling into the local optimum too early, and greatly improves the accuracy of optimization.

[0041] In some embodiments of the present invention, Figure 4 FIG. 1 is a flow chart of a random selection update strategy according to an embodiment of the present invention. Figure 4 As shown, a transition vector is generated by randomly selecting and updating the candidate solution balancing pool, including: S401, randomly selecting a first candidate solution from a candidate solution balancing pool, and randomly selecting a second candidate solution from an initial particle swarm; S402, randomly selecting a first candidate solution or a second candidate solution based on a preset selection probability, and optimizing particle positions of the particles to be updated in the initial particle swarm to obtain a transition vector based on a preset first position update strategy corresponding to the first candidate solution or a preset second position update strategy corresponding to the second candidate solution.

[0042] Specifically, in the balance optimizer algorithm, the original particle update formula is:

[0043] in, is the updated particle concentration, is the candidate solution in the balanced pool, is the current particle concentration, is the exponential term coefficient, is the generation rate, is a random vector, is the unit volume.

[0044] In order to improve the flexibility and search ability of the algorithm in the optimization process, the embodiment adopts a random selection update strategy in the particle population update, and obtains information from the balancing pool or random individuals with a preset probability to update the particle position, so that the algorithm can better adapt to different optimization problems. Specifically, the formula of the random selection update strategy is expressed as:

[0045] in, is a random candidate solution in the balancing pool, To randomly select individuals, is the transition vector, is the adaptive weight, for A random number, is the jump distance, is the adaptive weight, is the exponential term coefficient, is the generation rate, is a random vector, is the unit volume.

[0046] Example by random number Compared with the preset probability, the update method is randomly selected. When , the embodiment randomly selects a candidate solution from the balancing pool as a reference and calculates the transition vector through the corresponding formula of the first position update strategy. If , then the transition vector is calculated by randomly selecting a candidate solution from the particle population as a reference and using the corresponding formula of the second position update strategy. As an intermediate result obtained based on the random selection update strategy, the transition vector combines the candidate solution randomly selected from the balancing pool, the adaptive weight, and the information of other individuals in the population, provides a new candidate direction for updating the particle position, and decides whether to adopt the transition vector to update the particle through the subsequent Gaussian refresh strategy. In this way, the embodiment increases the randomness and diversity of particle updates, which helps the algorithm to jump out of the local optimum.

[0047] In some embodiments of the present invention, Figure 5 A schematic diagram of a Gaussian refresh strategy provided by an embodiment of the present invention is shown in FIG. Figure 5 As shown, according to the transition vector, the initial particle swarm is updated with the Gaussian refresh strategy population to obtain a new round of particle swarm, including: S501, determining the update fitness of the transition vector and the initial fitness of the particle to be updated according to the objective function; S502, if the updated fitness is greater than the initial fitness, the transition vector is used as a new round of particles in the particle swarm; S503: If the updated fitness is less than or equal to the initial fitness, the particles to be updated are subjected to Gaussian mutation update based on a preset mutation probability and are used as particles of a new round of the particle swarm.

[0048] Specifically, in order to help particles escape from local optimality and enhance global search capabilities, the embodiment adopts a Gaussian refresh strategy to update the population after randomly selecting an update strategy to obtain a transition vector. First, the embodiment determines whether to use a filter vector to update the position of the particle using the following formula:

[0049] in, Represents the transition vector The fitness of Indicates the particle to be updated When the particle can move to a better position , use the transition vector as a new round of particle update, when the particle cannot move to a better position , then according to the preset mutation probability, the particles are updated with Gaussian mutation to obtain a new round of particles. For particles that are not randomly mutated, the new round of particles remain unchanged. For particles that are randomly mutated and need Gaussian mutation, they are mutated according to the following formula:

[0050] In order to verify the effectiveness of this solution, the cold chain logistics path planning method of the present invention is experimented with by simulating path planning.

[0051] First, the embodiment sets the parameters of the improved balance optimization algorithm, sets the population size to N=80, the maximum number of algorithm iterations to T=150, and the number of algorithm runs to R=10 (this is used for averaging results from multiple runs, etc., and can be set according to actual needs).

[0052] Then, the adaptive weight strategy is used to initialize the initial position of the individuals in the population. The embodiment first determines the solution space range of the individuals in the population, that is, the value range of each node (customer point and distribution center) on the vehicle distribution path. The distribution center number is set to 0, and the customer point number ranges from 1 to n. The path is represented as a node sequence , where represents the customer point number, and each number appears only once in the path. For example, for n=5, a possible initial path is , the solution space range is all possible node sequence combinations. Then, the variable value is generated based on the adaptive weight formula, and the initial position of the individual in the population is determined by combining the upper and lower boundaries. exist The initial position of the first dimension is calculated as (in for random number between 1 and 2).

[0053] Then, the fitness value and pheromone concentration are generated according to the initial position of the individual (the pheromone concentration here can be compared to the historical quality of the path, which can be set to the same value at the beginning and updated later according to the quality of the path), and the fitness of the individual and the reverse solution are calculated through the objective function (total cost calculation function, such as in the agricultural product distribution scenario, the total cost includes fixed cost, transportation cost, refrigeration cost, carbon emission cost and penalty cost, etc.). Finally, the top half of the elite individual positions are selected according to the fitness and pheromone concentration.

[0054] The fitness value calculation process uses MATLAB and other software tools to model the distribution area and simulate the cost of vehicles in the distribution process. and unit distance cost To calculate, that is ; For cooling costs, the vehicle’s stay time at each node is calculated based on the delivery time window and driving time , according to the residence time and refrigeration unit cost Calculate, that is ; For carbon emission costs, the fuel consumption is determined based on the driving distance and vehicle fuel consumption rate , according to fuel consumption and carbon emissions per unit of fuel consumption Calculating Carbon Emissions , then multiply it by the carbon trading price to get ; For penalty cost, it is based on the time it takes for the vehicle to arrive at the customer point With the latest time If the relationship between > , then the penalty cost ,otherwise Total cost .

[0055] Finally, during the algorithm iteration process, the embodiment adopts a random selection update strategy to calculate the transition vector:

[0056] Then, according to the transition vector, the Gaussian refresh strategy population update is performed to update the population:

[0057] When particles can move to a better position , use the transition vector as a new round of particle update, when the particle cannot move to a better position , then according to the preset mutation probability, the particles are updated with Gaussian mutation to obtain a new round of particles. For particles that are not randomly mutated, the new round of particles remain unchanged. For particles that are randomly mutated and need Gaussian mutation, they are mutated according to the following formula:

[0058] Finally, the particle population is iteratively updated until the maximum number of iterations is reached to obtain the optimal path planning.

[0059] The optimization algorithm of this embodiment is experimentally compared with the genetic algorithm (GA) and the ant colony algorithm (ACA). The experimental results are statistically analyzed for the best value, worst value, average fitness value, and standard deviation of the running results. The specific experimental results are shown in Table 1 below: Table 1

[0060] As can be seen from the table, the improved balanced optimizer algorithm proposed in this embodiment has a relatively significant advantage in cold chain logistics path optimization. For example, in the F1 test case, the total cost of the optimal path solution obtained by the optimization algorithm (reflected by the inverse of the fitness value) is significantly lower than the results when the genetic algorithm and the ant colony algorithm are used alone. From the above results, it can be seen that the improved algorithm of this embodiment has a higher optimization ability, and the optimization results are better than other algorithms. Therefore, for the cold chain logistics path optimization problem, the optimization algorithm has better performance.

[0061] In addition, in order to verify that the algorithm of this embodiment can improve the efficiency of route planning in the cold chain logistics distribution area, this embodiment uses the MATLAB tool to simulate the cold chain logistics distribution area. First, a map model of the cold chain logistics distribution area is constructed, and different distribution scenarios are simulated by setting different parameters, such as setting 20 customer points, whose locations are randomly distributed in a certain area (longitude range [110,120], latitude range [30,40]), the distribution center is located at the center of the area (115,35), and the road network is assumed to be an undirected graph connecting all nodes (road weights can be set according to actual conditions, such as distance or driving time, etc.). Then, the starting point (distribution center location) and the end point (a certain customer point or a specific cargo distribution point, such as selecting the customer point farthest from the distribution center as the end point) are manually set in the distribution area, and finally a distribution area map is formed.

[0062] Figure 6 is a fitness iteration curve diagram of an embodiment of the present invention, such as Figure 6 As shown, this embodiment compares the optimization algorithm with the GA and ACA algorithms, where the parameter setting dimension is determined according to the actual delivery area, the population size is set as above, and the maximum number of iterations is. Other parameters are the same as above. Finally, Figure 6Fitness iteration curve diagram.

[0063] from Figure 6 It can be observed that the optimization algorithm of the present invention has a significant advantage in improving fitness. During the iteration process, the fitness improvement speed of the optimization algorithm of the present invention is significantly faster than that of the GA and ACA algorithms, and the fitness value is also significantly higher than that of the GA and ACA algorithms.

[0064] In order to better implement the cold chain logistics path planning method in the embodiment of the present invention, based on the cold chain logistics path planning method, correspondingly, Figure 7 As shown, the present invention also provides a cold chain logistics path planning device, and the cold chain logistics path planning device 700 includes: The planning condition building unit 701 is used to obtain the distribution related information and the variables to be planned for the cold chain logistics to be planned, and to build the objective function and constraint conditions of the cold chain logistics path planning according to the distribution related information and the variables to be planned; The iterative optimization solving unit 702 is used to initialize the variables to be planned to obtain an initial particle swarm, determine the particle fitness according to the objective function, construct a candidate solution balance pool according to the reverse learning strategy based on the particle fitness, generate a transition vector according to the random selection update strategy based on the candidate solution balance pool, update the population of the initial particle swarm according to the Gaussian refresh strategy according to the transition vector to obtain a new round of particle swarm, obtain the optimal particle position through the iterative update process, and determine the optimal cold chain logistics path according to the optimal particle position.

[0065] The cold chain logistics path planning device 700 provided in the above embodiment can implement the technical solution described in the above cold chain logistics path planning method embodiment. The specific implementation principles of the above modules or units can refer to the corresponding contents in the above cold chain logistics path planning method embodiment, which will not be repeated here.

[0066] like Figure 8 As shown, the present invention also provides an electronic device 800. The electronic device 800 includes a processor 801, a memory 802 and a display 803. Figure 8 Only some components of the electronic device 800 are shown, but it should be understood that it is not required to implement all of the components shown, and more or fewer components may be implemented instead.

[0067] In some embodiments, the processor 801 may be a central processing unit (CPU), a microprocessor or other data processing chip, used to run program codes or process data stored in the memory 802, such as the cold chain logistics path planning method of the present invention.

[0068] In some embodiments, the processor 801 may be a single server or a server group. The server group may be centralized or distributed. In some embodiments, the processor 801 may be local or remote. In some embodiments, the processor 801 may be implemented in a cloud platform. In one embodiment, the cloud platform may include a private cloud, a public cloud, a hybrid cloud, a community cloud, a distributed cloud, an internal cloud, a multi-cloud, etc., or any combination thereof.

[0069] In some embodiments, the memory 802 may be an internal storage unit of the electronic device 800, such as a hard disk or memory of the electronic device 800. In other embodiments, the memory 802 may also be an external storage device of the electronic device 800, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc., equipped on the electronic device 800.

[0070] Furthermore, the memory 802 may include both an internal storage unit of the electronic device 800 and an external storage device. The memory 802 is used to store application software installed in the electronic device 800 and various data.

[0071] In some embodiments, the display 803 may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, an OLED (Organic Light-Emitting Diode) touch device, etc. The display 803 is used to display information of the electronic device 800 and to display a visual user interface. The components 801-803 of the electronic device 800 communicate with each other via a system bus.

[0072] In one embodiment, when the processor 801 executes the cold chain logistics path planning program in the memory 802, the following steps may be implemented: Obtain the distribution-related information and variables to be planned for the cold chain logistics to be planned, and construct the objective function and constraint conditions for the cold chain logistics path planning based on the distribution-related information and variables to be planned; Initialize the variables to be planned to obtain the initial particle swarm, determine the particle fitness according to the objective function, use the reverse learning strategy to build the candidate solution equilibrium pool according to the particle fitness, generate the transition vector by randomly selecting the update strategy according to the candidate solution equilibrium pool, and update the initial particle swarm with the Gaussian refresh strategy according to the transition vector to obtain a new round of particle swarm. The optimal particle position is obtained through the iterative update process, and the optimal cold chain logistics path is determined according to the optimal particle position.

[0073] It should be understood that: when the processor 801 executes the cold chain logistics path planning program in the memory 802, in addition to the above functions, other functions can also be implemented. For details, please refer to the description of the corresponding method embodiment above.

[0074] Accordingly, an embodiment of the present application also provides a computer-readable storage medium, which is used to store computer-readable programs or instructions. When the program or instructions are executed by a processor, it can implement the steps or functions in the cold chain logistics path planning method provided in the above-mentioned method embodiments.

[0075] Those skilled in the art will appreciate that all or part of the processes of the above-mentioned embodiments can be implemented by instructing related hardware (such as a processor, a controller, etc.) through a computer program, and the computer program can be stored in a computer-readable storage medium, wherein the computer-readable storage medium is a disk, an optical disk, a read-only storage memory, or a random access memory, etc.

[0076] The cold chain logistics path planning method, device, electronic device and storage medium provided by the present invention are introduced in detail above. Specific examples are used in this article to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea; at the same time, for technical personnel in this field, according to the idea of ​​the present invention, there will be changes in the specific implementation method and application scope. In summary, the content of this specification should not be understood as a limitation on the present invention.

Claims

1. A cold chain logistics path planning method, characterized in that: include: Obtaining the distribution-related information and the variables to be planned for the cold chain logistics to be planned, and constructing the objective function and constraint conditions for the cold chain logistics path planning according to the distribution-related information and the variables to be planned; Initialize the variables to be planned to obtain an initial particle swarm, determine the particle fitness according to the objective function, perform a reverse learning strategy according to the particle fitness to construct a candidate solution balance pool, perform a random selection update strategy according to the candidate solution balance pool to generate a transition vector, perform a Gaussian refresh strategy population update on the initial particle swarm according to the transition vector to obtain a new round of particle swarm, obtain the optimal particle position through an iterative update process, and determine the optimal cold chain logistics path according to the optimal particle position.

2. The cold chain logistics path planning method according to claim 1, characterized in that: The delivery related information includes at least one of target location information, delivery demand, delivery time window requirement, initial delivery location and vehicle information, and the variables to be planned include at least one of delivery vehicle, delivery route and delivery start time.

3. The cold chain logistics path planning method according to claim 1, characterized in that: The objective function includes minimizing at least one of vehicle driving cost, cooling cost, carbon emission cost and penalty cost, and the constraint condition includes at least one of vehicle load constraint and time window constraint.

4. The cold chain logistics path planning method according to claim 1, characterized in that: Initializing the variables to be planned to obtain an initial particle swarm includes: Determine the particle search space boundary according to the constraint condition; Initial particles are constructed within the boundary of the particle search space based on preset adaptive weights to obtain an initial particle swarm.

5. The cold chain logistics path planning method according to claim 4, characterized in that: The method of performing a reverse learning strategy according to the particle fitness to construct a candidate solution balancing pool includes: Generate a mirror particle swarm according to the particle positions of the initial particle swarm and the boundary of the particle search space; After merging the initial particle group and the mirror particle group, fitness screening is performed to obtain several elite solutions, and a candidate solution balancing pool is constructed according to the elite solutions and the average value of the elite solutions.

6. The cold chain logistics path planning method according to claim 1, characterized in that: The step of generating a transition vector by randomly selecting an update strategy based on the candidate solution balancing pool includes: A first candidate solution is obtained by randomly selecting from the candidate solution balancing pool, and a second candidate solution is obtained by randomly selecting from the initial particle swarm; The first candidate solution or the second candidate solution is randomly selected based on a preset selection probability, and based on a preset first position update strategy corresponding to the first candidate solution or a preset second position update strategy corresponding to the second candidate solution, a particle position optimization is performed on the particles to be updated of the initial particle group to obtain a transition vector.

7. The cold chain logistics path planning method according to claim 6, characterized in that: The step of performing Gaussian refresh strategy population updating on the initial particle swarm according to the transition vector to obtain a new round of particle swarm includes: Determining the update fitness of the transition vector and the initial fitness of the particle to be updated according to the objective function; If the updated fitness is greater than the initial fitness, the transition vector is used as a new round of particles in the particle swarm; If the updated fitness is less than or equal to the initial fitness, the to-be-updated particle is updated by Gaussian mutation based on a preset mutation probability and then used as a new round of particles in the particle swarm.

8. A cold chain logistics path planning device, characterized in that: include: A planning condition construction unit, used to obtain the distribution-related information and the variables to be planned for the cold chain logistics to be planned, and to construct the objective function and constraint conditions for the cold chain logistics path planning according to the distribution-related information and the variables to be planned; The iterative optimization solving unit is used to initialize the variables to be planned to obtain an initial particle swarm, determine the particle fitness according to the objective function, perform a reverse learning strategy according to the particle fitness to construct a candidate solution balance pool, perform a random selection update strategy according to the candidate solution balance pool to generate a transition vector, perform a Gaussian refresh strategy population update on the initial particle swarm according to the transition vector to obtain a new round of particle swarm, obtain the optimal particle position through the iterative update process, and determine the optimal cold chain logistics path according to the optimal particle position.

9. An electronic device comprising a processor, a memory, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, it implements the cold chain logistics path planning method according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the cold chain logistics path planning method according to any one of claims 1 to 7 is implemented.

Images