Refrigerated vehicle distribution path optimization method based on evolutionary algorithm

Abstract

The invention discloses a refrigerated vehicle distribution path optimization method based on an evolutionary algorithm, and the method comprises the steps: 1, analyzing the demands of perishable goods for different transportation temperatures and the adaptation relation between different transportation temperature intervals of a heterogeneous refrigerated vehicle, and building a heterogeneous refrigerated vehicle distribution problem model; 2, two-stage optimization is carried out, in the first stage, a potential path of each refrigerator car is generated based on an evolutionary algorithm, and in the second stage, a comprehensive path scheme is generated through a path clustering and fusion mechanism; 3, iteratively searching a potential single path based on an evolutionary algorithm, regularly performing local search operation on the population to generate a potential path set, clustering the path set by using a clustering method, selecting paths in different clusters for multiple rounds for aggregation, and processing non-accessed clients to form a complete path scheme; and finally, outputting an optimal perishable goods refrigeration distribution scheme. According to the method, the problem of heterogeneous refrigeration vehicle distribution path optimization can be solved, and the goods loss is reduced to the maximum extent while the perishable goods distribution cost is reduced.

Description

Technical Field

[0001] The present invention relates to the field of cold chain transportation and vehicle route planning in the logistics industry, and specifically to a multi-temperature zone heterogeneous refrigerated vehicle route optimization method, which aims to improve the distribution efficiency of perishable goods and ensure the quality of the goods. Background Art

[0002] In the logistics industry, the efficient distribution of perishable goods is of vital importance and has attracted the attention of more and more researchers and practitioners who are committed to maximizing the supply of goods to the society. Researchers use various operations research tools to optimize the supply chain of goods and support the decision-making process. The transportation of perishable goods, including fresh fruits, vegetables, dairy products, and meat, requires precise temperature control in the supply chain to maintain their quality and safety. Therefore, the use of refrigerated vehicles has become an indispensable means to ensure that these goods remain fresh and edible during transportation.

[0003] In addition, with the newly released national standard, the types of mechanical refrigerated trucks used for the transportation of perishable goods are further subdivided (from Class A to Class I). According to the classification, the internal temperature range of refrigerated trucks is different, suitable for transporting various perishable goods with specific temperature requirements. This classification plays a key role in ensuring the safe transportation of sensitive goods.

[0004] In practice, transporting multiple products that require different storage temperatures simultaneously poses a significant challenge, especially when dealing with perishable goods that require different temperature ranges. For example, the temperature requirements of fresh produce and frozen goods vary greatly, making the joint transportation of these products more complicated. Different types of perishable goods may have conflicting environmental requirements, and if these special requirements cannot be met during transportation, it may lead to reduced freshness, deterioration, and even economic losses. When planning refrigerated fleets to transport a variety of perishable goods, existing route optimization strategies often fail to fully consider the diverse temperature requirements and characteristics of these goods, resulting in poor route planning and reduced cargo quality. Summary of the invention

[0005] In order to solve the problems existing in the above-mentioned prior art, the present invention proposes a refrigerated vehicle distribution path optimization method based on evolutionary algorithm, aiming to ensure efficient distribution route planning while meeting the temperature requirements of various perishable goods, thereby improving the distribution efficiency of perishable goods and reducing the economic losses and waste caused by quality decline.

[0006] The present invention adopts the following technical solutions to solve the above technical problems:

[0007] The invention discloses a refrigerated vehicle distribution path optimization method based on evolutionary algorithm, which is applied to a warehouse, customers, M types of perishable goods, and In the distribution service area composed of heterogeneous refrigerated trucks, The customer set consisting of customers is recorded as ; Among them, let The set of refrigerated trucks consisting of refrigerated trucks is denoted as V. The required temperature range of perishable goods is recorded as ,in, and Respectively indicate Minimum and maximum transport temperatures for perishable goods; The soft time window is denoted as ,in, and Respectively for customers The earliest start time and the latest end time of the service, ;

[0008] In the distribution service area, 1 warehouse and Each client acts as a node and builds a node set ; By node set The straight paths between nodes in the ;in, Represents any node and nodes The straight line path between The distance is ; Assuming that all refrigerated trucks have the warehouse as the starting point and the end point; then the method for optimizing the distribution path of refrigerated trucks for perishable goods is carried out in the following steps:

[0009] Step 1: Construct a perishable goods distribution model, including: objective function and constraints;

[0010] Step 1.1: Use equation (1) to establish the first objective function of the perishable goods distribution model :

[0011] (1)

[0012] In formula (1), Refrigerated truck stationary emissions; Refrigerated truck Whether to pass through the node and nodes The straight line path between , , Respectively represent refrigerated trucks Slave Node To Node The weight of the cargo carried, the length of the route and the transit time; The weight parameter representing the path length, represents the weight parameter of transportation time, Represents a constant parameter;

[0013] Step 1.2: Use formula (2) to establish the second objective function of the perishable goods distribution model :

[0014] (2)

[0015] In formula (2), Show to customers The weight of the goods delivered, It is a refrigerated truck from the warehouse to the customer The total time, Represents the refrigerated truck from the warehouse to the customer Number of previous deliveries; The weight parameter representing the number of previous deliveries, represents the weight parameter of delivery time, Represents a constant;

[0016] Step 1.3: Use formula (3) to establish the third objective function of the perishable goods distribution model :

[0017] (3)

[0018] In formula (3), To customers The earliest start time of the service, To customers The latest end time of the service; represents the weight parameter of the penalty for early arrival of vehicles, The weight parameter representing the penalty for vehicle lateness;

[0019] Step 1.4: Use formula (4) to establish the overall objective function of the perishable goods distribution model :

[0020] (4)

[0021] Step 1.5: Use equations (5) to (11) to construct the constraints of the perishable goods distribution model:

[0022] (5)

[0023] (6)

[0024] (7)

[0025] (8)

[0026] (9)

[0027] (10)

[0028] In formula (5), represents the maximum transport temperature of the refrigerated truck v, represents the minimum transport temperature of the refrigerated truck v, Indicates that, means or; ;

[0029] In formula (6), D represents the total demand of all customers. represents the total delivered cargo of refrigerated truck v;

[0030] In formula (7), Refrigerated truck Maximum load capacity;

[0031] In formula (8), Refrigerated truck Delivering perishable goods to customers Number of previous deliveries;

[0032] In formula (9), Indicates that it arrives from the warehouse to the customer The path length, represents the average speed of the refrigerated truck, Refrigerated truck At the Customer The uninstall time at the

[0033] Step 2: Use evolutionary algorithm to solve the perishable goods distribution model and obtain all the distribution path solutions of K types of refrigerated trucks;

[0034] Step 2.1: Define the current number of iterations ; Set the maximum number of iterations to ;

[0035] The type of refrigerated truck, the customer node served by the refrigerated truck, and the subscript code of the cargo type constitute a distribution route served by a single refrigerated truck and act as an individual. Individuals form a subpopulation of the tth generation, and thus the tth generation population P is composed of K subpopulations. t ={ },in, represents the kth subpopulation of the tth generation, and Only use Class K refrigerated trucks for delivery;

[0036] Step 2.2, initialize k=1;

[0037] Step 2.3, Initialization ;

[0038] Step 2.4: From the kth subpopulation of the tth generation Select two individuals in turn and perform crossover and mutation operations to generate two offspring individuals that meet the constraints, thereby obtaining N offspring individuals and Together they constitute the kth temporary subpopulation of the tth generation ;

[0039] Step 2.5: Calculate according to equations (1), (2) and (3): The objective function value of each individual in the perishable goods distribution model is used to After each individual in the non-dominated sort is sorted, the individuals of each level are selected in order from high to low according to the non-dominated level until the total number of individuals in all selected levels reaches N. If the total number of individuals exceeds N, the crowding distance is calculated for the individuals in the current level selected, so as to delete the individual with the smallest crowding distance, so that the total number of individuals in all levels is N, thus forming the kth subpopulation of the t+1th generation ;

[0040] Step 2.6: After assigning to t, if , then execute step 2.7, otherwise, return to step 2.4 and execute sequentially;

[0041] Step 2.7: After assigning k+1 to k, return to step 2.3 and execute sequentially until k>K, thus obtaining the first Generation population ={ },in, Indicates The kth subpopulation contains N delivery routes of kth type refrigerated trucks;

[0042] Step 3: Cluster the N delivery paths of K-type refrigerated trucks into a set of Y-type delivery paths:

[0043] Step 3.1: From Select the ath delivery path in turn and the bth delivery route , and the dynamic time warping method is used to calculate and The cumulative path distance between ;

[0044] Step 3.2: According to The cumulative distance between all delivery paths in the distribution path is clustered using the k-means algorithm. Clustering is performed to obtain a set of Y distribution paths { },in, represents the set of distribution paths after the yth clustering;

[0045] Step 4: Select a distribution path from each type of distribution path set after clustering and merge them into a complete refrigerated distribution plan for perishable goods, thereby obtaining a set of refrigerated distribution plans. ;

[0046] Step 4.1, define a perishable goods refrigerated distribution plan consisting of the types of refrigerated trucks, the customer nodes visited and served by the vehicles, and the subscript codes of the warehouse nodes as a distribution individual;

[0047] Step 4.2, define and initialize h=1;

[0048] Step 4.3, initialize y=1;

[0049] Step 4.4: From A delivery path r is randomly selected from the distribution path. If y=1, r is recorded as a temporary individual. Otherwise, get all customer nodes in r and compare them with the temporary individual Compare with the customer nodes in r and remove the nodes with After the repeated customer node, the removed delivery path r' is inserted into the temporary individual middle;

[0050] Step 4.5, after assigning y+1 to y, if , then the final temporary individual is obtained , and execute step 4.6, otherwise, return to step 4.4 and execute sequentially;

[0051] Step 4.6: Query the final temporary individual Is there any unserved client node in ? If so, use the greedy algorithm under the constraints of equations (5), (6), (7) and (10) to insert all unserved client nodes into the final temporary individual Thus, we get the hth distribution individual Stored in the cold storage distribution solution collection Pop;

[0052] Step 4.7, after assigning h+1 to h, if , then execute step 4.8, otherwise, return to step 4.3 and execute sequentially; where H represents The total number of individuals in

[0053] Step 5: From Delete the same optimal path planning solution from the above to obtain the optimal perishable goods refrigerated distribution set .

[0054] The method for optimizing the route of refrigerated vehicles for perishable goods distribution based on a two-stage optimization algorithm of the present invention is also characterized in that step 3.1 comprises:

[0055] Step 3.1.1, from Select the ath delivery path and the bth delivery path ,and Contains m nodes, Contains n nodes, calculate Each node and The Euclidean distance between each node in the , thus obtaining a distance matrix with dimension m×n ;Will The Ith node in The Euclidean distance between the Jth nodes in is recorded as (I,J);

[0056] Step 3.1.2: Calculate using formula (11) The Ith node in The cumulative distance between the Jth nodes in :

[0057] (11)

[0058] In formula (11), express The Ith node in The cumulative distance between the J-1th nodes in express The I-1th node and The cumulative distance between the Jth nodes in , express The I-1th node and The cumulative distance between the J-1th nodes in the , inf means infinity; when I=m and J=n, the cumulative distance As and The cumulative distance of the path .

[0059] Further, the step 3.2 includes:

[0060] Step 3.2.1, set the current number of iterations , and initialize ; Set the maximum number of iterations to ;

[0061] Set the total number of clusters to Y, The paths in the The Y-type delivery path set of iterations { }, where Indicates The set of y-th type of delivery paths in the iteration; and ={ },in, express The xth delivery path in The number of paths in ;

[0062] Step 3.2.2, initialize y=1;

[0063] Step 3.2.3, initialize x=1, flag=1, intermediate value Max=inf, where inf means infinity;

[0064] Step 3.2.4, calculation and The cumulative path distance of each delivery path in { }, thus obtaining Total accumulated distance ;in, express With 'Delivery routes The cumulative distance of the path; [1,X];

[0065] Step 3.2.5: If , then Assign to , assign x to flag; otherwise, execute step 3.2.6;

[0066] Step 3.2.6, assign x+1 to x. If x>X, execute step 3.2.7. Otherwise, return to step 3.2.4 and execute sequentially.

[0067] Step 3.2.7: Select the flag delivery route As the The yth cluster center of the iteration ;

[0068] Step 3.2.8, assign y+1 to y. If y>Y, then the first All cluster centers of the iteration { }; Otherwise, return to step 3.2.3 and execute sequentially;

[0069] Step 3.2.9, Calculation Any path in With all cluster centers { The cumulative distance between { }, and Divide into the cluster where the cluster center corresponding to the minimum cumulative distance is located, so that All paths in the The cluster center of the iteration is in the cluster;

[0070] Step 3.2.10: +1 assigned to Then return to step 3.2.2 and execute sequentially until > So as to obtain the final set of Y-type distribution paths { }.

[0071] An electronic device of the present invention includes a memory and a processor, wherein the memory is used to store a program that supports the processor to execute the refrigerated vehicle distribution path optimization method, and the processor is configured to execute the program stored in the memory.

[0072] The present invention provides a computer-readable storage medium, wherein a computer program is stored on the computer-readable storage medium, and the computer program executes the steps of the refrigerated vehicle distribution path optimization method when the computer program is executed by a processor.

[0073] Compared with the prior art, the present invention has the following beneficial effects:

[0074] 1. The present invention proposes an optimization method for a single path search algorithm and a path clustering fusion mechanism. In the first stage, an efficient individual path is generated by an evolutionary algorithm, laying the foundation for overall optimization; in the second stage, the path clustering and fusion mechanism further improves the efficiency and accuracy of path planning. This method can not only find the optimal transportation path, but also reduce transportation costs while meeting multiple temperature requirements.

[0075] 2. The present invention introduces a path grouping and optimization method based on the DTW (Dynamic Time Warping) algorithm and the K-means clustering algorithm. In the prior art, path optimization is usually based on simple distance calculations, while the present invention calculates the Euclidean distance between path nodes, constructs a cumulative distance matrix, and uses the DTW algorithm to effectively capture the dynamic changes between paths, making path grouping more accurate and reasonable. This method ensures the similarity between paths by minimizing the cumulative Euclidean distance, while greatly improving the accuracy of path clustering. BRIEF DESCRIPTION OF THE DRAWINGS

[0076] Figure 1 This is an example diagram of the dynamic time warping method of the present invention;

[0077] Figure 2 Flow chart of the method of the present invention. DETAILED DESCRIPTION

[0078] In this implementation example, a refrigerated truck delivery route optimization method based on an evolutionary algorithm is a two-stage optimization framework. Different stages focus on different aspects, which can better explore solutions at the boundary of the feasible domain. Under the condition of satisfying multiple strong constraints, it can improve the efficiency of searching for the minimum transportation cost, cargo quality loss and minimized waiting cost. Specifically, the refrigerated truck delivery route optimization method based on an evolutionary algorithm is applied to a warehouse consisting of 1 warehouse, customers, M types of perishable goods, and In the distribution service area composed of various heterogeneous refrigerated trucks, The customer set consisting of customers is recorded as ; Among them, let The set of refrigerated trucks consisting of refrigerated trucks is denoted as V. The required temperature range of perishable goods is recorded as ,in, and Respectively indicate Minimum and maximum transport temperatures for perishable goods; The soft time window is denoted as ,in, and Respectively for customers The earliest start time and the latest end time of the service, ;

[0079] In the distribution service area, 1 warehouse and Each client acts as a node and builds a node set ; By node set The straight paths between nodes in the ;in, Represents any node and nodes The straight line path between The distance is ; Assume that all refrigerated trucks start and end at the warehouse; refer to Figure 2 As shown in the figure, the refrigerated vehicle distribution path optimization method based on evolutionary algorithm is carried out in the following steps:

[0080] Step 1: Construct a perishable goods distribution model, including: objective function and constraints;

[0081] Step 1.1: Use equation (1) to establish the first objective function of the perishable goods distribution model :

[0082] (1)

[0083] In formula (1), Refrigerated truck stationary emissions; Refrigerated truck Whether to pass through the node and nodes The straight line path between , , Respectively represent refrigerated trucks Slave Node To Node The weight of the cargo carried, the length of the route and the transit time; The weight parameter representing the path length, represents the weight parameter of transportation time, Represents a constant parameter.

[0084] Step 1.2: Use formula (2) to establish the second objective function of the perishable goods distribution model :

[0085] (2)

[0086] In formula (2), Show to customers The weight of the goods delivered, It is a refrigerated truck from the warehouse to the customer The total time, Represents the refrigerated truck from the warehouse to the customer Number of previous deliveries; The weight parameter representing the number of previous deliveries, represents the weight parameter of delivery time, Represents a constant.

[0087] Step 1.3: Use formula (3) to establish the third objective function of the perishable goods distribution model :

[0088] (3)

[0089] In formula (3), To customers The earliest start time of the service, To customers The latest end time of the service; represents the weight parameter of the penalty for early arrival of vehicles, The weight parameter representing the penalty for vehicle lateness;

[0090] Step 1.4: Use formula (4) to establish the overall objective function of the perishable goods distribution model :

[0091] (4).

[0092] Step 1.5: Use equations (5) to (11) to construct the constraints of the perishable goods distribution model:

[0093] (5)

[0094] (6)

[0095] (7)

[0096] (8)

[0097] (9)

[0098] (10)

[0099] In formula (5), represents the maximum transport temperature of the refrigerated truck v, represents the minimum transport temperature of the refrigerated truck v, Indicates that, means or; ;

[0100] In formula (6), D represents the total demand of all customers. represents the total delivered cargo of refrigerated truck v;

[0101] In formula (7), Refrigerated truck Maximum load capacity;

[0102] In formula (8), Refrigerated truck Delivering perishable goods to customers Number of previous deliveries;

[0103] In formula (9), Indicates that it arrives from the warehouse to the customer The path length, represents the average speed of the refrigerated truck, Refrigerated truck At the Customer The uninstall time at the time.

[0104] For example, the demand information of 5 customers is recorded in Table 1 and the temperature range information of each type of refrigerated truck is recorded in Table 2. In this embodiment, there are 7 types of refrigerated trucks, and the maximum load capacity of each type of refrigerated truck is 25. The transportation temperature can be adjusted arbitrarily within its own temperature zone.

[0105] Table 1 Customer requirements

[0106]

[0107] Table 2 Temperature range information of refrigerated trucks

[0108]

[0109] Step 2: Use evolutionary algorithm to solve the perishable goods distribution model and obtain all the distribution path solutions of K types of refrigerated trucks;

[0110] Step 2.1: Define the current number of iterations ; Set the maximum number of iterations to ;

[0111] The type of refrigerated truck, the customer node served by the refrigerated truck, and the subscript code of the cargo type constitute a distribution route served by a single refrigerated truck and act as an individual. Individuals form a subpopulation of the tth generation, and thus the tth generation population P is composed of K subpopulations. t ={ },in, represents the kth subpopulation of the tth generation, and Only use Class K refrigerated trucks for delivery;

[0112] For example, the decision variables of the five individuals in the Class A refrigerated truck population are initialized as [0,A0223,A0312,A0428,A0503,0], [0,A0312,A0428,A0223,A0503,0], [0,A0428,A0223,A0503,A 0312,0], [0,A0503,A0223,A0312,A0428, 0], [0,A0223,A0312,A0503,A 0428,0], where the first letter indicates the type of refrigerated truck, the first two digits indicate the customer subscript and the last two digits indicate the customer's cargo number. For example, A0223 means that a Class A refrigerated truck delivers Class 23 cargo for customer No. 2.

[0113] Step 2.2, initialize k=1;

[0114] Step 2.3, Initialization ;

[0115] Step 2.4: From the kth subpopulation of the tth generation Select two individuals in turn and perform crossover and mutation operations to generate two offspring individuals that meet the constraints, thereby obtaining N offspring individuals and Together they constitute the kth temporary subpopulation of the tth generation ;

[0116] Step 2.5: Calculate according to equations (1), (2) and (3): The objective function value of each individual in the perishable goods distribution model is shown in Figure 1. Taking the first individual as an example, its first objective value is the distribution cost of the Class A refrigerated truck, which includes fixed usage cost, driving cost, compartment refrigeration cost, and related costs of loading and unloading perishable goods. Substituting into the formula, the route cost is 173.6. The second objective is the quality loss of the delivered goods. Substituting into the formula value, the cargo quality loss value is 13.2. The second objective is the waiting time cost. If the vehicle meets the customer's time window, the cost is 0. If it arrives earlier than the time window, the refrigerated truck will wait for the customer, incurring additional costs. If it arrives late, customer satisfaction will be greatly reduced. Substituting into formula (3), the waiting cost is calculated to be 3.5. The remaining four individuals are calculated to obtain Table 3:

[0117] Table 3:

[0118]

[0119] Used for After each individual in the non-dominated sort is sorted, the individuals of each level are selected in order from high to low according to the non-dominated level until the total number of individuals in all selected levels reaches N. If the total number of individuals exceeds N, the crowding distance is calculated for the individuals in the current level selected, so as to delete the individual with the smallest crowding distance, so that the total number of individuals in all levels is N, thus forming the kth subpopulation of the t+1th generation .

[0120] Step 2.6: After assigning to t, if , then execute step 2.7, otherwise, return to step 2.4 and execute sequentially;

[0121] Step 2.7: After assigning k+1 to k, return to step 2.3 and execute sequentially until k>K, thus obtaining the first Generation population ={ },in, Indicates The kth subpopulation contains N delivery routes of the kth type of refrigerated trucks.

[0122] Step 3: Cluster the N delivery paths of K-type refrigerated trucks into a set of Y-type delivery paths:

[0123] Step 3.1: From Select the ath delivery path in turn and the bth delivery route , and the dynamic time warping method is used to calculate and The cumulative path distance between ;

[0124] Step 3.1.1, from Select the ath delivery path and the bth delivery path ,and Contains m nodes, Contains n nodes, calculate Each node and The Euclidean distance between each node in the , thus obtaining a distance matrix with dimension m×n ;Will The Ith node in The Euclidean distance between the Jth nodes in is recorded as (I,J);

[0125] Step 3.1.2: Calculate using formula (11) The Ith node in The cumulative distance between the Jth nodes in :

[0126] (11)

[0127] In formula (11), express The Ith node in The cumulative distance between the J-1th nodes in express The I-1th node and The cumulative distance between the Jth nodes in , express The I-1th node and The cumulative distance between the J-1th nodes in the , inf means infinity; when I=m and J=n, the cumulative distance As and The cumulative distance of the path .

[0128] For example, Figure 1 As shown, there are two delivery routes ={0,3,6,4}, ={0,2,5}, Table 4 is the distance matrix Record the distance between each customer in the two paths; build and The cumulative distance matrix , calculated by formula (11) and the distance matrix The value of each position in the result is shown in Table 5. (4,3) = 5.6 is the measure and Cumulative distance between paths ;

[0129] Table 4 Distance matrix

[0130]

[0131] Table 5 Cumulative distance matrix

[0132]

[0133] Step 3.2: According to The cumulative distance between all delivery paths in the distribution path is clustered using the k-means algorithm. Clustering is performed to obtain a set of Y distribution paths { },in, represents the set of distribution paths after the yth clustering;

[0134] Step 3.2.1, set the current number of iterations , and initialize ; Set the maximum number of iterations to ;

[0135] Set the total number of clusters to Y, The paths in the The Y-type delivery path set of iterations { }, where Indicates The set of y-th type of delivery paths in the iteration; and ={ },in, express The xth delivery path in The number of paths in .

[0136] Step 3.2.2, initialize y=1;

[0137] Step 3.2.3, initialize x=1, flag=1, intermediate value Max=inf, where inf means infinity;

[0138] Step 3.2.4, calculation and The cumulative path distance of each delivery path in { }, thus obtaining Total accumulated distance ;in, express With 'Delivery routes The cumulative distance of the path; [1,X];

[0139] Step 3.2.5: If , then Assign to , assign x to flag; otherwise, execute step 3.2.6;

[0140] Step 3.2.6, assign x+1 to x. If x>X, execute step 3.2.7. Otherwise, return to step 3.2.4 and execute sequentially.

[0141] Step 3.2.7: Select the flag delivery route As the The yth cluster center of the iteration ;

[0142] Step 3.2.8, assign y+1 to y. If y>Y, then the first All cluster centers of the iteration { }; Otherwise, return to step 3.2.3 and execute sequentially.

[0143] Step 3.2.9, Calculation Any path in With all cluster centers { The cumulative distance between { }, and Divide into the cluster where the cluster center corresponding to the minimum cumulative distance is located, so that All paths in the The cluster center of the iteration is in the cluster;

[0144] Step 3.2.10: +1 assigned to Then return to step 3.2.2 and execute sequentially until > So as to obtain the final set of Y-type distribution paths { }.

[0145] Step 4: Select a distribution path from each type of distribution path set after clustering and merge them into a complete refrigerated distribution plan for perishable goods, thereby obtaining a set of refrigerated distribution plans. ;

[0146] Step 4.1, define a perishable goods refrigerated distribution plan consisting of the types of refrigerated trucks, the customer nodes visited and served by the vehicles, and the subscript codes of the warehouse nodes as a distribution individual;

[0147] Step 4.2, define and initialize h=1;

[0148] Step 4.3, initialize y=1;

[0149] Step 4.4: From A delivery path r is randomly selected from the distribution path. If y=1, r is recorded as a temporary individual. Otherwise, get all customer nodes in r and compare them with the temporary individual Compare with the customer nodes in r and remove the nodes with After the repeated customer node, the removed delivery path r' is inserted into the temporary individual middle;

[0150] Step 4.5, after assigning y+1 to y, if , then the final temporary individual is obtained , and execute step 4.6, otherwise, return to step 4.4 and execute sequentially;

[0151] Step 4.6: Query the final temporary individual Is there any unserved client node in ? If so, use the greedy algorithm under the constraints of equations (5), (6), (7) and (10) to insert all unserved client nodes into the final temporary individual Thus, we get the hth distribution individual Stored in the cold storage distribution solution collection Pop;

[0152] Step 4.7, after assigning h+1 to h, if , then execute step 4.8, otherwise, return to step 4.3 and execute sequentially; where H represents The total number of individuals in

[0153] Step 5: From Delete the same optimal path planning solution from the above to obtain the optimal perishable goods refrigerated distribution set .

[0154] In this embodiment, an electronic device includes a memory and a processor, wherein the memory is used to store a program that supports the processor to execute the above method, and the processor is configured to execute the program stored in the memory.

[0155] In this embodiment, a computer-readable storage medium stores a computer program on the computer-readable storage medium, and the computer program executes the steps of the above method when executed by a processor.

Claims

1. A refrigerated vehicle distribution route optimization method based on evolutionary algorithm, characterized by applying 1 warehouse, customers, M types of perishable goods, and In the distribution service area composed of various heterogeneous refrigerated trucks, The customer set consisting of customers is recorded as ; Among them, let The set of refrigerated trucks consisting of refrigerated trucks is denoted as V. The required temperature range of perishable goods is recorded as ,in, and Respectively indicate The minimum and maximum transportation temperatures for the delivery of perishable goods; The soft time window is denoted as ,in, and Respectively for customers The earliest start time and the latest end time of the service, ; In the distribution service area, one warehouse and Each client acts as a node and builds a node set ; By node set The straight paths between nodes in the ;in, Represents any node and nodes The straight line path between The distance is ; Assuming that all refrigerated trucks have the warehouse as the starting point and the end point; then the method for optimizing the distribution path of refrigerated trucks for perishable goods is carried out in the following steps: Step 1: Construct a perishable goods distribution model, including: objective function and constraints; Step 1.1: Use equation (1) to establish the first objective function of the perishable goods distribution model : (1) In formula (1), Refrigerated truck stationary emissions; Refrigerated truck Whether to pass through the node and nodes The straight line path between , , Respectively represent refrigerated trucks Slave Node To Node The weight of the cargo carried, the length of the route and the transit time; The weight parameter representing the path length, represents the weight parameter of transportation time, Represents a constant parameter; Step 1.2: Use formula (2) to establish the second objective function of the perishable goods distribution model : (2) In formula (2), Show to customers The weight of the goods delivered, It is a refrigerated truck from the warehouse to the customer The total time, Represents the refrigerated truck from the warehouse to the customer Number of previous deliveries; The weight parameter representing the number of previous deliveries, represents the weight parameter of delivery time, Represents a constant; Step 1.3: Use formula (3) to establish the third objective function of the perishable goods distribution model : (3) In formula (3), To customers The earliest start time of the service, To customers The latest end time of the service; represents the weight parameter of the penalty for early arrival of vehicles, The weight parameter representing the penalty for vehicle lateness; Step 1.4: Use formula (4) to establish the overall objective function of the perishable goods distribution model : (4) Step 1.5: Use equations (5) to (11) to construct the constraints of the perishable goods distribution model: (5) (6) (7) (8) (9) (10) In formula (5), represents the maximum transport temperature of the refrigerated truck v, represents the minimum transport temperature of the refrigerated truck v, Indicates that, means or; ; In formula (6), D represents the total demand of all customers. represents the total delivered cargo of refrigerated truck v; In formula (7), Refrigerated truck Maximum load capacity; In formula (8), Refrigerated truck Delivering perishable goods to customers Number of previous deliveries; In formula (9), Indicates that it arrives from the warehouse to the customer The path length, represents the average speed of the refrigerated truck, Refrigerated truck At the Customer The uninstall time at the Step 2: Use evolutionary algorithm to solve the perishable goods distribution model and obtain all the distribution path solutions of K types of refrigerated trucks; Step 2.1: Define the current number of iterations ; Set the maximum number of iterations to ; The type of refrigerated truck, the customer node served by the refrigerated truck, and the subscript code of the cargo type constitute a distribution route served by a single refrigerated truck and act as an individual. Individuals form a subpopulation of the tth generation, and thus the tth generation population P is composed of K subpopulations. t ={ },in, represents the kth subpopulation of the tth generation, and Only use Class K refrigerated trucks for delivery; Step 2.2, initialize k=1; Step 2.3, Initialization ; Step 2.4: From the kth subpopulation of the tth generation Select two individuals in turn and perform crossover and mutation operations to generate two offspring individuals that meet the constraints, thereby obtaining N offspring individuals and Together they constitute the kth temporary subpopulation of the tth generation ; Step 2.5: Calculate according to equations (1), (2) and (3): The objective function value of each individual in the perishable goods distribution model is used to After each individual in the non-dominated sort is sorted, the individuals of each level are selected in order from high to low according to the non-dominated level until the total number of individuals in all selected levels reaches N. If the total number of individuals exceeds N, the crowding distance of the individuals in the current level is calculated, and the individual with the smallest crowding distance is deleted, so that the total number of individuals in all levels is N, thus forming the kth subpopulation of the t+1th generation ; Step 2.6: After assigning to t, if , then execute step 2.7, otherwise, return to step 2.4 and execute sequentially; Step 2.7: After assigning k+1 to k, return to step 2.3 and execute sequentially until k>K, thus obtaining the first Generation population ={ },in, Indicates The kth subpopulation contains N delivery routes of kth type refrigerated trucks; Step 3: Cluster the N delivery paths of K-type refrigerated trucks into a set of Y-type delivery paths: Step 3.1: From Select the ath delivery path in turn and the bth delivery route , and the dynamic time warping method is used to calculate and The cumulative path distance between ; Step 3.2: According to The cumulative distance between all delivery paths in the distribution path is clustered using the k-means algorithm. Clustering is performed to obtain a set of Y distribution paths { },in, represents the set of distribution paths after the yth clustering; Step 4: Select a distribution path from each type of distribution path set after clustering and merge them into a complete refrigerated distribution plan for perishable goods, thereby obtaining a set of refrigerated distribution plans. ; Step 4.1, define a perishable goods refrigerated distribution plan consisting of the types of refrigerated trucks, the customer nodes visited and served by the vehicles, and the subscript codes of the warehouse nodes as a distribution individual; Step 4.2, define and initialize h=1; Step 4.3, initialize y=1; Step 4.4: From A delivery path r is randomly selected from the distribution path. If y=1, r is recorded as a temporary individual. Otherwise, get all customer nodes in r and compare them with the temporary individual Compare with the customer nodes in r and remove the nodes with After the repeated customer node, the removed delivery path r' is inserted into the temporary individual middle; Step 4.5, after assigning y+1 to y, if , then the final temporary individual is obtained , and execute step 4.6, otherwise, return to step 4.4 and execute sequentially; Step 4.6: Query the final temporary individual Is there any unserved client node in ? If so, use the greedy algorithm under the constraints of equations (5), (6), (7) and (10) to insert all unserved client nodes into the final temporary individual Thus, we get the hth distribution individual Stored in the cold storage distribution solution collection Pop; Step 4.7, after assigning h+1 to h, if , then execute step 4.8, otherwise, return to step 4.3 and execute sequentially; where H represents The total number of individuals in Step 5: From Delete the same optimal path planning solution from the above to obtain the optimal perishable goods refrigerated distribution set .

2. The method for optimizing the route of refrigerated vehicles for perishable goods distribution based on a two-stage optimization algorithm according to claim 1 is characterized in that: The step 3.1 comprises: Step 3.1.1, from Select the ath delivery path and the bth delivery path ,and Contains m nodes, Contains n nodes, calculate Each node and The Euclidean distance between each node in the , thus obtaining a distance matrix with dimension m×n ;Will The Ith node in The Euclidean distance between the Jth nodes in is recorded as (I, J); Step 3.1.2: Calculate using formula (11) The Ith node in The cumulative distance between the Jth nodes in : (11) In formula (11), express The Ith node in The cumulative distance between the J-1th nodes in , express The I-1th node in The cumulative distance between the Jth nodes in , express The I-1th node and The cumulative distance between the J-1th nodes in the , inf means infinity; when I=m and J=n, the cumulative distance As and The cumulative distance of the path .

3. The method for optimizing the route of refrigerated vehicles for perishable goods distribution based on a two-stage optimization algorithm according to claim 1 is characterized in that: The step 3.2 comprises: Step 3.2.1, set the current number of iterations , and initialize ; Set the maximum number of iterations to ; Set the total number of clusters to Y, The paths in the The Y-type delivery path set of iterations { }, where Indicates The set of y-th type of delivery paths in the iteration; and ={ },in, express The xth delivery path in The number of paths in ; Step 3.2.2, initialize y=1; Step 3.2.3, initialize x=1, flag=1, intermediate value Max=inf, where inf means infinity; Step 3.2.4, calculation and The cumulative path distance of each delivery path in { }, thus obtaining Total accumulated distance ;in, express With 'Delivery routes The cumulative distance of the path; [1,X]; Step 3.2.5: If , then Assign to , assign x to flag; otherwise, execute step 3.2.6; Step 3.2.6, assign x+1 to x. If x>X, execute step 3.2.

7. Otherwise, return to step 3.2.4 and execute sequentially. Step 3.2.7, select the flag delivery route As the The yth cluster center of the iteration ; Step 3.2.8, assign y+1 to y. If y>Y, then the first All cluster centers of the iteration { }; Otherwise, return to step 3.2.3 and execute sequentially; Step 3.2.9, Calculation Any path in With all cluster centers { The cumulative distance between { }, and Divide into the cluster where the cluster center corresponding to the minimum cumulative distance is located, so that All paths in the The cluster center of the iteration is in the cluster; Step 3.2.10: +1 assigned to Then return to step 3.2.2 and execute sequentially until > So as to obtain the final set of Y-type distribution paths { }.

4. An electronic device, comprising a memory and a processor, characterized in that: The memory is used to store a program that supports the processor to execute the refrigerated vehicle distribution path optimization method described in any one of claims 1-3, and the processor is configured to execute the program stored in the memory.

5. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the refrigerated vehicle delivery route optimization method described in any one of claims 1 to 3 are executed.

Images