Multi-farm picking-distribution integrated scheduling method and system

Abstract

The invention relates to a multi-farm picking-distribution integrated scheduling method and system, and relates to the field of agricultural product supply chain management and optimization, and the method comprises the steps: setting a constraint condition for a multi-farm picking-distribution integrated scheduling problem; defining two objective functions, including an operation cost minimization function and a customer satisfaction maximization function; based on the constraint condition and the objective function, establishing a dual-objective model of multi-farm picking-distribution integrated scheduling; and solving the dual-objective optimization model to obtain an optimal solution set, and realizing picking and distribution of agricultural products according to picking and distribution information in the optimal solution set. Picking and distribution links of multiple farms are effectively integrated, and the freshness of agricultural products is effectively guaranteed.

Description

Technical Field

[0001] This invention relates to the field of agricultural product supply chain management and optimization technology, and in particular to a multi-farm harvesting-distribution integrated scheduling method and system. Background Technology

[0002] In recent years, with the continuous improvement of living standards, people have paid increasing attention to healthy eating, leading to a surge in demand for fresh and organic agricultural products. As an emerging agricultural product supply chain model, the "direct sourcing and supply" model refers to farmers establishing direct contact with end consumers through electronic platforms, shortening distribution links, and aiming to provide consumers with fresh and healthy agricultural products. Under this operational framework, farms provide a one-stop service from "harvesting to delivery," ensuring that fresh agricultural products can be harvested, delivered, and arrived on the same day.

[0003] However, while the "direct sourcing and supply" model has developed rapidly, it also faces challenges such as the difficulty of preserving agricultural products and high harvesting costs. Because farms are typically located on the outskirts of cities and customers are widely distributed, delivery efficiency is low and delivery costs are high. Furthermore, the freshness of fresh agricultural products is a key factor determining their value; once harvested, their freshness gradually decreases over time. In addition, to meet diverse consumer demands, multiple farms often cooperate and operate jointly. Therefore, how to effectively integrate the harvesting and delivery processes of multiple farms to provide high-freshness agricultural products at low cost has become a critical issue that urgently needs to be addressed.

[0004] However, current research largely focuses on a single stage of harvesting or distribution, with insufficient exploration of comprehensive optimization for these two stages. Furthermore, research integrating the harvesting and distribution processes of multiple heterogeneous farms into a unified scheduling framework is extremely rare. Summary of the Invention

[0005] Therefore, the technical problem to be solved by the present invention is to overcome the problems of insufficient integration of agricultural product harvesting and distribution processes, high costs, and difficulty in guaranteeing freshness in the existing "direct procurement and supply" model.

[0006] To address the aforementioned technical problems, this invention provides a multi-farm harvesting-delivery integrated scheduling method, comprising:

[0007] To address the integrated scheduling problem of multi-farm harvesting and delivery, constraints are set, including: each farm grows a variety of agricultural products, and the types of agricultural products grown on each farm are completely different, all of which are provided to customers; the harvesting team is arranged to harvest agricultural products in an optimized harvesting order according to customer needs, and vehicles are arranged to depart from the depot to the farm to load agricultural products, then deliver the agricultural products to customers, and finally return to the depot.

[0008] Define two objective functions: minimizing operating costs and maximizing customer satisfaction. Operating costs include harvesting costs, transportation costs, and vehicle usage costs. Customer satisfaction is related to the freshness of the delivered agricultural products and the delivery time.

[0009] Based on the constraints and the objective function, a dual-objective model for integrated scheduling of multi-farm harvesting and delivery is established.

[0010] Solve the bi-objective optimization model to obtain the optimal solution set, and realize the harvesting and distribution of agricultural products based on the harvesting and distribution information in the optimal solution set.

[0011] In one embodiment of the present invention, the constraint conditions further include:

[0012] Each type of agricultural product can only be harvested by one harvesting team;

[0013] Each picking team can only pick one type of agricultural product at a time;

[0014] The harvesting efficiency of the picking team is not constant;

[0015] Each type of agricultural product has a specific rate of freshness loss;

[0016] Each customer needs at least one agricultural product;

[0017] Each customer is delivered by only one vehicle;

[0018] The specific farm to which the vehicle will go will be determined based on the type of product required by the customer.

[0019] The time it takes for the vehicle to leave the farm is equal to the time it takes for the last type of agricultural product to be loaded to be harvested.

[0020] The total load of vehicles at any node on the route shall not exceed the vehicle's carrying capacity.

[0021] All vehicles must depart from the parking lot and return to the parking lot after serving the customer.

[0022] In one embodiment of the present invention, the function for minimizing operating costs is expressed as:

[0023] ;

[0024] In the formula, Represents operating costs, This represents the harvesting cost per unit of time. This represents the cost of travel per unit of time. This represents the fixed operating cost of each vehicle; Representative Team Picking Farm agricultural products Total harvesting time; Represents the vehicle from the node To the node Travel time; If the vehicle Access Node Then visit the node ,but ;otherwise, ; , represents the node index, where , representing a set of nodes , representing a collection of farms Represents the number of farms, , representing a collection of customers Represents the total number of customers. Represents the parking lot; , representing the vehicle index, where , representing a collection of vehicles Represents the number of vehicles; , representing the farm index; , representing the index of the picking team, in which This represents the gathering of the picking team. Representative Farm The number of picking teams in China; , representing an agricultural product index, in which , representing a collection of agricultural products, where 0 represents virtual agricultural products. Representative Farm The variety and quantity of agricultural products in China; If the vehicle From the parking lot to the farm ,but ;otherwise, .

[0025] In one embodiment of the invention, the function for maximizing customer satisfaction is represented by minimizing freshness loss and delivery delay:

[0026] ;

[0027] In the formula, This represents loss of freshness and delivery delays; This represents the customer's dissatisfaction with the product's freshness. On behalf of clients To capture the maximum loss of freshness in agricultural products; This represents the customer's dissatisfaction with the product delivery time. Representative visits clients The resulting delivery delays; , , representing a collection of customers Represents the number of farms, Represents the total number of customers.

[0028] In one embodiment of the present invention, the bi-objective model of the multi-farm harvesting-delivery integrated scheduling is expressed as follows:

[0029] ;

[0030] ;

[0031] ;

[0032] ;

[0033] ;

[0034] ;

[0035] ;

[0036] ;

[0037] ;

[0038] ;

[0039] ;

[0040] ;

[0041] ;

[0042] ;

[0043] ;

[0044] ;

[0045] ;

[0046] ;

[0047] ;

[0048] ;

[0049] ;

[0050] ;

[0051] ;

[0052] ;

[0053] ;

[0054] ;

[0055] ;

[0056] ;

[0057] ;

[0058] ;

[0059] ;

[0060] ;

[0061] ;

[0062] ;

[0063] ;

[0064] ;

[0065] ;

[0066] In the formula, , representing an agricultural product index, in which , representing a collection of agricultural products, where 0 represents virtual agricultural products. Representative Farm The variety and quantity of agricultural products in China; This represents a set of agricultural products that do not contain 0. , represents the node index, where , representing a set of nodes , representing a collection of farms Represents the number of farms, , representing a collection of customers Represents the total number of customers. Representing the car park, , representing the vehicle index, where , representing a collection of vehicles Represents the number of vehicles; , representing the index of the picking team, in which This represents the gathering of the picking team. Representative Farm The number of picking teams in China; On behalf of clients Total demand; If the customer Farm needed agricultural products but ;otherwise, ; It is a farm agricultural products Freshness loss rate; Represents the vehicle from the node To the node Travel time; This represents the maximum load capacity of each vehicle; On behalf of clients The latest required delivery time; If the farm The picking team After harvesting agricultural products Then came the harvesting of agricultural products. ,but ;otherwise, ; If the farm The picking team After harvesting agricultural products Then came the harvesting of agricultural products. ,but ;otherwise, ; Representative Farm The picking team Picking virtual products Then came the harvesting of agricultural products. ,but ;otherwise, ; If the farm The picking team After harvesting agricultural products Then they picked agricultural products. ,but ;otherwise, ; If the vehicle Serving customers ,but ;otherwise, ; Representative vehicle leave Time; Representative vehicle Arrival at the farm Time; Representative picking farm agricultural products The relative start time; Representative Farm agricultural products The time for completion of harvesting; For farm agricultural products The time for completion of harvesting; If the farm The picking team Picking agricultural products ,but ;otherwise, ; If the farm The picking team Picking agricultural products ,but ;otherwise, ; Representing clients on farms Products in Total demand; Represents infinity; Representative of the picking team For the farm Chinese agricultural products The actual harvesting efficiency; Representative of the picking team For the farm Chinese agricultural products Standard harvesting efficiency; Represents the dynamic harvesting efficiency coefficient; This represents the duration during which the harvesting team maintained peak efficiency without experiencing fatigue; This represents the fatigue duration of the harvesting team; This represents the entire cycle of the harvesting team's efficiency, from peak to decline due to fatigue, and then back to rest and recovery. Representative Farm Chinese agricultural products Delivered to customer The loss of freshness that occurs over time; If the vehicle From the parking lot Heading to the farm ,but Otherwise, ; If the vehicle Access Node Then return to the parking lot ,but Otherwise, ; If the vehicle From the parking lot Head to customer location ,but ;otherwise, ; If the vehicle From the farm Heading to the parking lot ,but ;otherwise, ; If the vehicle Visit customer point Then we went to the farm ,but ;otherwise, ; If the vehicle Access Node Then visit the node ,but ;otherwise, ; Representative vehicle Leaving the parking lot Time; Representative vehicle Leaving the parking lot Time; If the vehicle Serving customers ,but ;otherwise, ; If the vehicle Access Node Then visit the node ,but ; If the vehicle Access Node Then visit the node ,but ;otherwise, ; Indicates if vehicle Visit the farm ,but ;otherwise, ; Representative car field to the farm Travel time; Representative Farm to the farm Travel time; If the vehicle From the farm Go to node ,but Otherwise, ; If the vehicle From the farm Heading to the farm ,but Otherwise, ; Representative vehicle Arrival at the farm Time; Representative vehicle Arrival at the farm Time; Representative vehicles from the farm To the customer point Travel time; Represents the vehicle from the node To the node Travel time; If the vehicle From the farm Head to customer location ,but Otherwise, ; If the vehicle From the customer point Head to customer location ,but Otherwise, ; This indicates that the vehicle has arrived at the customer's location. Time; The vehicle arrived at the customer's location. Time; Representative Team Picking Farm medium-sized products Total harvesting time; Representative Farm Chinese agricultural products Freshness loss rate; On behalf of clients To capture the maximum loss of freshness in agricultural products; Indicates customer The resulting delivery delays.

[0067] In one embodiment of the present invention, the dynamic harvesting efficiency function Represented as:

[0068] ;

[0069] In the formula, The time for picking; The duration for which the harvesting team can maintain peak efficiency without experiencing fatigue; This represents the peak efficiency stage for the harvesting team. To reduce the fatigue rate of the harvesting team; The fatigue duration of the harvesting team; This represents the fatigue stage of the harvesting team; The duration of the entire cycle for the harvesting team, from peak efficiency to decreased efficiency due to fatigue, and then to rest and recovery of efficiency. This represents the rest period for the picking team.

[0070] In one embodiment of the present invention, the method for solving the bi-objective optimization model to obtain the optimal solution set includes:

[0071] Two initial subpopulations are randomly generated: a picking subpopulation and a delivery subpopulation. Each subpopulation contains... An individual represented by an integer encoding;

[0072] Merging individuals with the same index from two subpopulations yields A complete solution, evaluation Two objective functions are used to obtain a set of optimal solutions as the optimal solution set, which is then stored in an external file.

[0073] The current harvested seed population and the delivery seed population are input into the trained DQN model, and the DQN model outputs the evolutionary strategy of the current harvested seed population and the delivery seed population. The evolutionary strategy includes three types: crossover and mutation, and crossover and mutation.

[0074] The two subpopulations after implementing the evolutionary strategy will be used to update the optimal solution set in the external archive.

[0075] Determine if the maximum running time has been reached. If it has, terminate the process and output the optimal solution set. Otherwise, continue to input the two subpopulations with the updated evolutionary strategy into the trained DQN model until the maximum running time is reached.

[0076] In one embodiment of the present invention, the picking subpopulation is encoded with integers, including the task allocation of the picking team and the order of agricultural product picking;

[0077] The delivery subpopulation is encoded with integers, including the order of farm and customer visits;

[0078] The crossover includes uniform crossover, single-point crossover, and sequential crossover. The uniform crossover is only used for the integer corresponding to the task allocation of the picking team in the picking sub-population. The single-point crossover is only used for the integer corresponding to the picking order of agricultural products in the picking sub-population. The sequential crossover is only used for the integer corresponding to the customer index in the delivery sub-population.

[0079] The mutation is applied only to the integers corresponding to the picking order of agricultural products in the picking subpopulation and the integers corresponding to the customer index in the delivery subpopulation, and the mutation is to swap the positions of the two arbitrary integers.

[0080] To solve the above-mentioned technical problems, the present invention provides a multi-farm harvesting-delivery integrated scheduling system, comprising:

[0081] The configuration module is used to set constraints for the integrated scheduling problem of multi-farm harvesting and delivery, including: each farm grows a variety of agricultural products, and the types of agricultural products grown on each farm are completely different, and they all provide agricultural products to customers; according to customer needs, the harvesting team is arranged to harvest agricultural products in an optimized harvesting order, and vehicles are arranged to depart from the depot to the farm to load agricultural products, deliver agricultural products to customers, and finally return to the depot.

[0082] Definition module: Used to define two objective functions, including minimizing operating costs and maximizing customer satisfaction. Operating costs include harvesting costs, transportation costs, and vehicle usage costs, while customer satisfaction is related to the freshness of the delivered agricultural products and the delivery time.

[0083] Establishment Module: Used to establish a dual-objective model for integrated scheduling of multi-farm harvesting and delivery based on the constraints and the objective function;

[0084] Solution module: Used to solve the bi-objective optimization model to obtain the optimal solution set, and realize the harvesting and distribution of agricultural products based on the harvesting and distribution information in the optimal solution set.

[0085] To address the aforementioned technical problems, the present invention provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the steps of the multi-farm harvesting-delivery integrated scheduling method described above.

[0086] To address the aforementioned technical problems, the present invention provides a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the steps of the multi-farm harvesting-delivery integrated scheduling method described above.

[0087] Compared with the prior art, the above-described technical solution of the present invention has the following advantages:

[0088] The multi-farm harvesting-delivery integrated scheduling method described in this invention involves multiple farms growing completely different agricultural products, cooperating with each other to provide agricultural products to customers. The method integrates the two key links of harvesting and delivery. By introducing an operational model of collaborative cooperation among multiple heterogeneous farms and implementing a two-stage integrated scheduling strategy for multiple farms, the freshness of agricultural products is effectively guaranteed.

[0089] This invention designs a dual-population optimization algorithm to solve a dual-objective model. The algorithm is carefully designed in terms of encoding rules, crossover strategy, and local search method, and incorporates deep reinforcement learning technology to improve the algorithm's solution efficiency. Attached Figure Description

[0090] To make the content of this invention easier to understand, the invention will be further described in detail below with reference to specific embodiments and accompanying drawings.

[0091] Figure 1 This is a flowchart of the multi-farm harvesting-delivery integrated scheduling method of the present invention;

[0092] Figure 2 This is a flowchart of the dual-population optimization algorithm in an embodiment of the present invention;

[0093] Figure 3 This is a schematic diagram of the encoding and decoding of the solution in an embodiment of the present invention;

[0094] Figure 4 These are schematic diagrams of the uniform cross method and the single-point cross method in embodiments of the present invention;

[0095] Figure 5 This is a schematic diagram of the sequential crossing method in an embodiment of the present invention;

[0096] Figure 6 This is a box plot of the IGD indexes obtained by the four algorithms in different computational examples in this embodiment of the invention;

[0097] Figure 7 This is a box plot of the HV index obtained by the four algorithms in different computational examples in the embodiments of the present invention. Detailed Implementation

[0098] The present invention will be further described below with reference to the accompanying drawings and specific embodiments, so that those skilled in the art can better understand and implement the present invention. However, the embodiments described are not intended to limit the present invention.

[0099] Example 1

[0100] Reference Figure 1 As shown, this invention proposes a multi-farm harvesting-delivery integrated scheduling method, including:

[0101] To address the integrated scheduling problem of multi-farm harvesting and delivery, constraints are set, including: each farm grows multiple agricultural products, and the types of agricultural products grown by each farm are completely different (for example, if one farm grows cabbage and radishes, then other farms cannot grow cabbage and radishes), and they cooperate to provide agricultural products to customers; based on the received customer orders, customer needs are obtained, and harvesting tasks and vehicle delivery tasks are arranged for each farm according to customer needs, that is, the harvesting team is arranged to harvest agricultural products in an optimized harvesting order, and vehicles are arranged to depart from the depot to the farm to load agricultural products, deliver agricultural products to customers, and finally return to the depot;

[0102] Define two objective functions: minimizing operating costs and maximizing customer satisfaction. Operating costs include harvesting costs, transportation costs, and vehicle usage costs. Customer satisfaction is related to the freshness of the delivered agricultural products and the delivery time.

[0103] Based on constraints and objective functions, a dual-objective model for integrated scheduling of multi-farm harvesting and delivery is established.

[0104] Solve the bi-objective optimization model to obtain the optimal solution set, and carry out the harvesting and distribution of agricultural products based on the harvesting and distribution information in the optimal solution set.

[0105] As can be seen from the above technical solution, this embodiment proposes a multi-farm harvesting-delivery integrated scheduling method, aiming to solve the problems of insufficient integration, high costs, and difficulty in guaranteeing freshness in the existing "direct sourcing and supply" model for agricultural product harvesting and delivery. This method integrates the two major links of harvesting and delivery, and introduces collaborative cooperation among multiple heterogeneous farms, achieving full-chain optimization of agricultural products from harvesting to delivery, providing consumers with diversified products. Simultaneously, this embodiment defines an agricultural product freshness loss function, a dynamic harvesting efficiency function, and two objective functions, constructing a dual-objective model that includes minimizing operating costs and maximizing customer satisfaction. By solving this model, this invention obtains a Pareto optimal solution, achieving dual optimization of operating costs and customer satisfaction. Furthermore, this embodiment proposes a dual-population optimization algorithm, with careful design in encoding, crossover strategies, and local search, and incorporates deep reinforcement learning technology, significantly improving solution efficiency. This method not only reduces operating costs and minimizes agricultural product freshness loss but also improves service satisfaction, providing practical guidance and useful reference for farmers and the fresh food e-commerce industry, demonstrating strong practical application value and superiority.

[0106] The multi-farm harvesting and delivery integrated scheduling problem of the present invention includes four decisions: (1) the task allocation of harvesting teams in each farm, that is, the allocation of harvested agricultural products to harvesting teams; (2) the harvesting order of agricultural products of harvesting teams; (3) the allocation between customer points and transport vehicles, that is, the allocation of customer points to transport vehicles; and (4) the route of vehicles.

[0107] Specifically, for the integrated scheduling problem of multi-farm harvesting and delivery, specific constraints are set, including: multiple farms growing completely different agricultural products cooperate to provide agricultural products to customers; upon receiving customer orders, harvesting tasks and vehicle delivery tasks are arranged for each farm according to customer needs; that is, harvesting teams are arranged to harvest agricultural products in an optimized order, and vehicles are arranged to depart from the depot to the farms to load agricultural products, deliver agricultural products to customers, and finally return to the depot; each farm grows multiple agricultural products, and the types of agricultural products grown on each farm are completely different; each agricultural product can only be harvested by one harvesting team; each harvesting team can only harvest one agricultural product at a time; the harvesting efficiency of the harvesting teams is not constant; each agricultural product has a specific freshness loss rate (i.e., The freshness of agricultural products decreases over time after harvesting; each customer may need more than one type of agricultural product; each customer is delivered by only one vehicle; the vehicle to serve the customer is determined based on the types of products required by the customer; the time when the vehicle leaves the farm is equal to the time when the last type of agricultural product to be loaded is harvested; the vehicle's carrying capacity is limited, that is, the total load of vehicles at any point on the route does not exceed the vehicle's carrying capacity; all vehicles must depart from the depot and return to the depot after serving the customer.

[0108] The dynamic harvesting efficiency function in this embodiment reflects the fatigue effect and rest and recovery strategies of the harvesting team, and is expressed as follows:

[0109] ;

[0110] In the formula, This represents the actual harvesting efficiency of the harvesting team. The time for picking; The duration for which the harvesting team can maintain peak efficiency without experiencing fatigue; This represents the highest efficiency range of the harvesting team, where the harvesting efficiency is the highest, which is 1 (i.e., 100%). The fatigue level of the harvesting team indicates a decrease in their harvesting efficiency index. The duration of the entire cycle for the harvesting team, from peak efficiency to decreased efficiency due to fatigue, and then to rest and recovery of efficiency. The picking efficiency of the team in the rest area recovers linearly.

[0111] It should be noted that the above and The following factors affect the calculation results of the bi-objective optimization model.

[0112] Specifically, this embodiment defines two objective functions: minimizing operating costs and maximizing customer satisfaction. Operating costs are a key objective of the model optimization, encompassing harvesting costs, transportation costs, and vehicle usage costs. Harvesting costs are influenced by the efficiency of the harvesting team, while transportation and vehicle usage costs are related to delivery routes and vehicle allocation. Therefore, by optimizing the harvesting team's efficiency and delivery routes, operating costs can be reduced. Specifically, the function for minimizing operating costs can be expressed as:

[0113] ;

[0114] In the formula, Represents operating costs, This represents the harvesting cost per unit of time. This represents the cost of travel per unit of time. This represents the fixed operating cost of each vehicle; Representative Team Picking Farm agricultural products Total harvesting time; Represents the vehicle from the node To the node Travel time; If the vehicle Access Node Then visit the node ,but ;otherwise, ; , represents the node index, where , representing a set of nodes , representing a collection of farms Represents the number of farms, , representing a collection of customers Represents the total number of customers. Represents the parking lot; , representing the vehicle index, where , representing a collection of vehicles Represents the number of vehicles; , representing the farm index; , representing the index of the picking team, in which This represents the gathering of the picking team. Representative Farm The number of picking teams in China; , representing an agricultural product index, in which , representing a collection of agricultural products, where 0 represents virtual agricultural products. Representative Farm The variety and quantity of agricultural products in China; If the vehicle From the parking lot to the farm ,but ;otherwise, .

[0115] Customer satisfaction is another important optimization objective, related to the freshness and delivery time of the delivered agricultural products. Higher freshness and more punctual delivery lead to higher customer satisfaction. The freshness of agricultural products is influenced by a freshness loss function. Specifically, maximizing customer satisfaction is represented by minimizing freshness loss and delivery delay:

[0116] ;

[0117] In the formula, This represents loss of freshness and delivery delays; This represents the customer's dissatisfaction with the product's freshness. On behalf of clients To capture the maximum loss of freshness in agricultural products; This represents the customer's dissatisfaction with the product delivery time. Representative visits clients The resulting delivery delays; , , representing a collection of customers Represents the number of farms, Represents the total number of customers.

[0118] Specifically, based on the above constraints and objective function, this embodiment establishes a dual-objective model for integrated scheduling of multi-farm harvesting and delivery, wherein the dual-objective optimization model is expressed as follows:

[0119] ;

[0120] ;

[0121] ;

[0122] ;

[0123] ;

[0124] ;

[0125] ;

[0126] ;

[0127] ;

[0128] ；

[0129] ；

[0130] ；

[0131] ；

[0132] ；

[0133] ；

[0134] ；

[0135] ；

[0136] ；

[0137] ；

[0138] ；

[0139] ；

[0140] ；

[0141] ；

[0142] ；

[0143] ；

[0144] ；

[0145] ；

[0146] ；

[0147] ；

[0148] ；

[0149] ；

[0150] ；

[0151] ；

[0152] ；

[0153] ;

[0154] ;

[0155] ;

[0156] In the formula, , representing an agricultural product index, in which , represents the set of agricultural products, where 0 represents virtual agricultural products (the purpose of setting virtual agricultural products in this embodiment is to achieve the constraint closed loop of the model). Representative Farm The variety and quantity of agricultural products in China; This represents a set of agricultural products that do not contain 0. , represents the node index, where , representing a set of nodes , representing a collection of farms Represents the number of farms, , representing a collection of customers Represents the total number of customers. Representing the car park, , representing the vehicle index, where , representing a collection of vehicles Represents the number of vehicles; , representing the index of the picking team, in which This represents the gathering of the picking team. Representative Farm The number of picking teams in China; On behalf of clients Total demand; If the customer Farm needed agricultural products but ;otherwise, ; It is a farm agricultural products Freshness loss rate; Represents the vehicle from the node To the node Travel time; This represents the maximum load capacity of each vehicle; On behalf of clients The latest required delivery time; If the farm The picking team After harvesting agricultural products Then came the harvesting of agricultural products. ,but ;otherwise, ; If the farm The picking team After harvesting agricultural products Then came the harvesting of agricultural products. ,but ;otherwise, ; Representative Farm The picking team Picking virtual products Then came the harvesting of agricultural products. ,but ;otherwise, ; If the farm The picking team After harvesting agricultural products Then they picked agricultural products. ,but ;otherwise, ; If the vehicle Serving customers ,but ;otherwise, ; Representative vehicle leave Time; Representative vehicle Arrival at the farm Time; Representative picking farm agricultural products The relative start time; Representative Farm agricultural products The time for completion of harvesting; For farm agricultural products The time for completion of harvesting; If the farm The picking team Picking agricultural products ,but ;otherwise, ; If the farm The picking team Picking agricultural products ,but ;otherwise, ; Representing clients on farms Products in Total demand; Represents infinity; Representative of the picking team For the farm Chinese agricultural products The actual harvesting efficiency; Representative of the picking team For the farm Chinese agricultural products Standard harvesting efficiency; Represents the dynamic harvesting efficiency coefficient; This represents the duration during which the harvesting team maintained peak efficiency without experiencing fatigue; This represents the fatigue duration of the harvesting team; This represents the entire cycle of the harvesting team's efficiency, from peak to decline due to fatigue, and then back to rest and recovery. Representative Farm Chinese agricultural products Delivered to customer The loss of freshness that occurs over time; If the vehicle From the parking lot Heading to the farm ,but Otherwise, ; If the vehicle Access Node Then return to the parking lot ,but Otherwise, ; If the vehicle From the parking lot Head to customer location ,but ;otherwise, ; If the vehicle From the farm Heading to the parking lot ,but ;otherwise, ; If the vehicle Visit customer point Then we went to the farm ,but ;otherwise, ; If the vehicle Access Node Then visit the node ,but ;otherwise, ; Representative vehicle Leaving the parking lot Time; Representative vehicle Leaving the parking lot Time; If the vehicle Serving customers ,but ;otherwise, ; If the vehicle Access Node Then visit the node ,but ; If the vehicle Access Node Then visit the node ,but ;otherwise, ; Indicates if vehicle Visit the farm ,but ;otherwise, ; Representative car field to the farm Travel time; Representative Farm to the farm Travel time; If the vehicle From the farm Go to node ,but Otherwise, ; If the vehicle From the farm Heading to the farm ,but Otherwise, ; Representative vehicle Arrival at the farm Time; Representative vehicle Arrival at the farm Time; Representative vehicles from the farm To the customer point Travel time; Represents the vehicle from the node To the node Travel time; If the vehicle From the farm Head to customer location ,but Otherwise, ; If the vehicle From the customer point Head to customer location ,but Otherwise, ; This indicates that the vehicle has arrived at the customer's location. Time; The vehicle arrived at the customer's location. Time; Representative Team Picking Farm medium-sized products Total harvesting time; Representative Farm Chinese agricultural products Freshness loss rate; On behalf of clients To capture the maximum loss of freshness in agricultural products; Indicates customer The resulting delivery delays.

[0157] Specifically, this embodiment uses an improved dual-population optimization algorithm to solve the multi-objective optimization model, obtaining the optimal solution set (Pareto equilibrium solution). Based on the harvesting and delivery information in the optimal solution set, agricultural product scheduling is performed, including determining customer order allocation, harvesting task allocation, harvesting order, vehicle allocation, and delivery routes. Specific steps include:

[0158] Step S1: Initialize algorithm parameters, including population size, mutation rate, maximum running time, etc.;

[0159] Step S2: Randomly generate two initial subpopulations (harvesting subpopulation and delivery subpopulation), each subpopulation containing An individual represented by an integer encoding;

[0160] Step S3: Obtain the result by merging individuals with the same index from the two subpopulations. A complete solution (see details) Figure 2 ),Evaluate The objective values ​​of each solution (i.e., minimizing the operating cost function and maximizing the customer satisfaction function mentioned above) are used to obtain a set of optimal solutions as the optimal solution set (the optimal solution set may have 1, 2, 3, etc.), and the optimal solution set is stored in an external file.

[0161] Step S4: Input the current picking subpopulation and delivery subpopulation into the trained DQN model, and output the evolutionary strategy of the current picking subpopulation and delivery subpopulation through the DQN model (the two subpopulations execute the same evolutionary strategy). The evolutionary strategy includes three types: crossover and mutation, and crossover and mutation.

[0162] Step S5: Update the optimal solution set in the external archive with the two subpopulations after implementing the evolutionary strategy;

[0163] The specific update method in step S5 is as follows: calculate the minimum operating cost function and the maximum customer satisfaction function for the two subpopulations after implementing the evolutionary strategy, and calculate the minimum operating cost function and the maximum customer satisfaction function for each solution in the optimal solution set in the external archive. Individuals with the same index in the two subpopulations after implementing the evolutionary strategy are grouped into a new... A complete solution, for this The complete solution operation is as follows: Each complete solution is compared one by one with each solution in the optimal solution set. If a solution in the optimal solution set is dominated by the complete solution (i.e., both objective functions of the complete solution are better), then that solution is removed from the optimal solution set, and the complete solution is copied to the optimal solution set (this is the process of checking this). If a complete solution can dominate the solutions in the optimal solution set, then remove the dominated optimal solution and copy the better complete solution into it; otherwise, keep the optimal solution set unchanged. This process is repeated continuously, requiring the addition of new solutions. All complete solutions have been compared with the optimal solution set.

[0164] Step S6: Determine if the maximum running time has been reached. If it has, terminate and output the optimal solution set; otherwise, return to step S4 and continue to input the two subpopulations with the updated evolutionary strategy into the trained DQN model until the maximum running time is reached.

[0165] Furthermore, this embodiment also uses integer encoding to represent the feasible solution to the problem, including two integer strings, integer string 1 ( ) represents the harvesting decision: the task allocation of the harvesting team and the order of agricultural product harvesting, and the integer string 2 ( This indicates the delivery decision: the order in which the farm and the customer are visited. For example... Figure 3 As shown, from integer string 1, we know that: Farm 1's harvesting team 1 is responsible for harvesting two kinds of agricultural products, harvesting agricultural product 3 and agricultural product 1 in sequence; Farm 1's harvesting team 2 is responsible for harvesting one kind of agricultural product, namely agricultural product 2; Farm 2's harvesting team 1 is responsible for harvesting one kind of agricultural product, harvesting agricultural product 4 in sequence; Farm 2's harvesting team 2 is responsible for harvesting three kinds of agricultural products, namely harvesting agricultural product 7, agricultural product 6, and agricultural product 5 in sequence. From integer string 2, we know that: Vehicle 1 departs from the parking lot, goes to Farm 2 and Farm 1 to load agricultural products, then goes to customers 8, 6, 7, and 4 in sequence to deliver products, and finally returns to the parking lot; Vehicle 2 departs from the parking lot, goes to Farm 1 and Farm 2 to load agricultural products, then goes to customers 3, 9, and 5 in sequence to deliver products, and finally returns to the parking lot. Note: Harvesting integer string Individuals forming subpopulation 1; distributing integer strings Individuals that make up subpopulation 2.

[0166] In this embodiment, the mutation is an exchange-type mutation, and the mutation is applied to the integer corresponding to the agricultural product picking order in subpopulation 1 (picking subpopulation) and to the integer corresponding to the customer index in subpopulation 2 (delivery subpopulation). For example Figure 3 In the example, if you randomly select farm 1, the integers corresponding to the harvesting order of its agricultural products are 3, 1, and 2. Mutation involves randomly swapping the positions of any two numbers from 3, 1, and 2. For example... Figure 3 In the process, if route 1 is chosen arbitrarily, the integers corresponding to its customer index are 8, 6, 7, and 4. Mutation is to randomly swap the positions of any two numbers among 8, 6, 7, and 4.

[0167] The intersections in this embodiment include uniform intersections, single-point intersections, and sequential intersections. For example... Figure 4 and Figure 5 The diagram illustrates the crossover operation between subpopulation 1 (picking subpopulation) and subpopulation 2 (delivery subpopulation). In the process of finding the optimal solution set using a dual-population optimization algorithm, subpopulation 1 uses both uniform crossover and single-point crossover methods to obtain new individuals, while subpopulation 2 uses sequential crossover. The uniform crossover (only used for the integers corresponding to the task assignments of the picking teams in subpopulation 1) is designed as follows: a random number 0 or 1 is generated; if the number 1 is generated, then the new individuals are selected from parent generation 1 (…). Figure 4 The gene is replicated from generation 2113. If a number 0 is generated, it is copied from generation 2 (in the parent generation). Figure 4 In the 1222 section, the duplicated gene (parent 1 and parent 2 are two parent individuals selected from the subpopulation using a binary selection method, specifically selecting individuals with a better objective function). The single-point crossover (only used for integers corresponding to the agricultural product harvesting order in subpopulation 1) is designed as follows: For parent 1 (… Figure 4Randomly generate a cut point (between 3 and 1, and between 7 and 6) for Farm 1 (corresponding to 312, and Farm 2 (corresponding to 4765)). Select the gene segment to the left of the cut point (312 corresponds to 3, and 4765 corresponds to 47) and directly copy it to the corresponding gene position of Offspring 1 (initial Offspring). In Parent Generation 2 (231, and 5467), delete the copied genes (i.e., 3 and 47) of Parent Generation 1. The remaining genes (2 and 1 corresponding to Farm 1, and 5 and 6 corresponding to Farm 2) are assigned to Offspring 1 in sequence. This implementation also repairs the offspring 1 (initial offspring), that is, it repairs (the repair is only used for uniform crossover) the integers corresponding to the task assignments of the harvesting teams generated by the uniform crossover in the initial offspring. These integers must be the same as the sum of the integers at the positions of parent generation 1 / parent generation 2. In the initial offspring (i.e., 22321234756), the sum of 2 and 2 (the integers corresponding to the task assignments of the harvesting teams) for farm 1 is 4 (exceeding the 3 of parent generation 1 / parent generation 2), and the sum of 2 and 3 (the integers corresponding to the task assignments of the harvesting teams) for farm 2 is 5 (exceeding the 4 of parent generation 1 / parent generation 2). Therefore, in this embodiment, each repair of the initial offspring involves randomly subtracting 1 from one of the integers 2 and 2 corresponding to farm 1. Figure 4 The process involves subtracting the first 2 from the 2 and 2 corresponding to farm 1, resulting in the repaired 1 and 2 corresponding to farm 1 (at this point, the sum is 3, the same as the two parent generations). The repair of the initial offspring involves subtracting 1 from a random integer between 2 and 3 corresponding to farm 2, resulting in the repaired 1 and 3 corresponding to farm 1 (at this point, the sum is 4, the same as the two parent generations). If the first round of repair does not meet the conditions, the loop continues until the conditions are met, ultimately completing the repair of the initial offspring and obtaining the final offspring. For example... Figure 4 The sequence 12321134756. The sequential crossover (only used for integers corresponding to customer indices in subpopulation 2) is designed as follows: Randomly select parent generation 1 and parent generation 2 of subpopulation 2 (specifically, select individuals with better objective functions), first concatenate the integers corresponding to customer indices from different routes in parent generation 1 and parent generation 2 (e.g., ...). Figure 5 In this process, the client index 8674 of route 1 and the client index 395 of route 2 in parent generation 1 are concatenated to obtain 8674395, and the client index 37965 of route 1 and the client index 48 of route 2 in parent generation 2 are concatenated to obtain 3796548. The concatenated lengths of the two parent generations are the same. Two different cut points are randomly selected in the parent generation (the first cut point is between 6 and 7 in 8674395, and the second cut point is between 3 and 9). Gene segments between the cut points are selected. Figure 5 In this example, 743 is the crossover segment. The crossover segment (743) from parent 1 is copied to the corresponding gene locus in the offspring. Genes (7, 4, 3) from parent 2 that cross over with parent 1 are deleted. The remaining genes (9, 6, 5, 8) are assigned to the offspring in sequence, resulting in 9674358. It should be noted that... Figure 5 Assuming that the agricultural products corresponding to customer index 967 are loaded to the maximum capacity of the vehicle, and cannot be loaded further, then the agricultural products corresponding to customer index 4358 will be loaded onto other vehicles. Figure 5 Assuming that the agricultural products needed by customer index 967 are all in farm 1, then 1 is set before 967. Figure 5 Assuming that the agricultural products required by customer index 4358 are in farm 1 and farm 2, then 4358 is initially set to 1 and 2, and based on route optimization, farm 1 and 2 are eventually set to 21, resulting in offspring 1 (1967214358).

[0168] To verify the effectiveness and superiority of the dual-population optimization algorithm (DQN-MCC) proposed in this embodiment, three algorithms were selected as comparison algorithms: Non-dominated sorting genetic algorithm II (NSGA-II), decomposition-based multi-objective evolutionary algorithm (MOEA / D), and Q-learning-assisted co-evolutionary algorithm (HVQ-MCC). These algorithms were run 20 times on examples of different sizes. The IGD and HV metrics were used to measure the algorithm performance. It is worth noting that a smaller IGD value indicates better algorithm performance, while a larger HV value indicates better algorithm performance. Figure 6 and Figure 7 Box plots of the four algorithms on the IGD and HV indices are presented respectively. It can be seen that the dual-population optimization algorithm in this embodiment performs better than the other three algorithms.

[0169] Example 2

[0170] This embodiment provides a multi-farm harvesting-delivery integrated scheduling system, including:

[0171] The configuration module is used to set constraints for the integrated scheduling problem of multi-farm harvesting and delivery, including: each farm grows a variety of agricultural products, and the types of agricultural products grown on each farm are completely different, and they all provide agricultural products to customers; according to customer needs, the harvesting team is arranged to harvest agricultural products in an optimized harvesting order, and vehicles are arranged to depart from the depot to the farm to load agricultural products, deliver agricultural products to customers, and finally return to the depot.

[0172] Definition module: Used to define two objective functions, including minimizing operating costs and maximizing customer satisfaction. Operating costs include harvesting costs, transportation costs, and vehicle usage costs, while customer satisfaction is related to the freshness of the delivered agricultural products and the delivery time.

[0173] Establishment Module: Used to establish a dual-objective model for integrated scheduling of multi-farm harvesting and delivery based on the constraints and the objective function;

[0174] Solution module: Used to solve the bi-objective optimization model to obtain the optimal solution set, and realize the harvesting and distribution of agricultural products based on the harvesting and distribution information in the optimal solution set.

[0175] Example 3

[0176] This embodiment provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the steps of the multi-farm harvesting-delivery integrated scheduling method described in Embodiment 1.

[0177] Example 4

[0178] This embodiment provides a computer-readable storage medium storing a computer program thereon. When the computer program is executed by a processor, it implements the steps of the multi-farm harvesting-delivery integrated scheduling method described in Embodiment 1.

[0179] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code. The solutions in the embodiments of this application can be implemented in various computer languages, such as the object-oriented programming language Java and the interpreted scripting language JavaScript.

[0180] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0181] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0182] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0183] Obviously, the above embodiments are merely illustrative examples for clear explanation and are not intended to limit the implementation. Those skilled in the art will recognize that other variations or modifications can be made based on the above description. It is neither necessary nor possible to exhaustively list all possible implementations here. However, obvious variations or modifications derived therefrom are still within the scope of protection of this invention.

Claims

1. A multi-farm harvesting-delivery integrated scheduling method, characterized in that, include: To address the integrated scheduling problem of multi-farm harvesting and delivery, constraints are set, including: each farm grows a variety of agricultural products, and the types of agricultural products grown on each farm are completely different, all of which are provided to customers; the harvesting team is arranged to harvest agricultural products in an optimized harvesting order according to customer needs, and vehicles are arranged to depart from the depot to the farm to load agricultural products, then deliver the agricultural products to customers, and finally return to the depot. Define two objective functions: minimizing operating costs and maximizing customer satisfaction. Operating costs include harvesting costs, transportation costs, and vehicle usage costs. Customer satisfaction is related to the freshness of the delivered agricultural products and the delivery time. Based on the constraints and the objective function, a dual-objective model for integrated scheduling of multi-farm harvesting and delivery is established. Solve the bi-objective optimization model to obtain the optimal solution set, and realize the harvesting and distribution of agricultural products based on the harvesting and distribution information in the optimal solution set.

2. The multi-farm harvesting-delivery integrated scheduling method according to claim 1, characterized in that: The constraints also include: Each type of agricultural product can only be harvested by one harvesting team; Each picking team can only pick one type of agricultural product at a time; The harvesting efficiency of the picking team is not constant; Each type of agricultural product has a specific rate of freshness loss; Each customer needs at least one agricultural product; Each customer is delivered by only one vehicle; The specific farm to which the vehicle will go will be determined based on the type of product required by the customer. The time it takes for the vehicle to leave the farm is equal to the time it takes for the last type of agricultural product to be loaded to be harvested. The total load of vehicles at any node on the route shall not exceed the vehicle's carrying capacity. All vehicles must depart from the parking lot and return to the parking lot after serving the customer.

3. The multi-farm harvesting-delivery integrated scheduling method according to claim 1, characterized in that: The function for minimizing operating costs is expressed as: ； In the formula, Represents operating costs, This represents the harvesting cost per unit of time. This represents the cost of travel per unit of time. This represents the fixed operating cost of each vehicle; Representative Team Picking Farm agricultural products Total harvesting time; Represents the vehicle from the node To the node Travel time; If the vehicle Access Node Then visit the node ,but ; otherwise, ; , represents the node index, where , representing a set of nodes , representing a collection of farms Represents the number of farms, , representing a collection of customers Represents the total number of customers. Represents the parking lot; , representing the vehicle index, where , representing a collection of vehicles Represents the number of vehicles; , representing the farm index; , representing the index of the picking team, in which This represents the gathering of the picking team. Representative Farm The number of picking teams in China; , representing an agricultural product index, in which , representing a collection of agricultural products, where 0 represents virtual agricultural products. Representative Farm The variety and quantity of agricultural products in China; If the vehicle From the parking lot to the farm ,but ; otherwise, .

4. The multi-farm harvesting-delivery integrated scheduling method according to claim 1, characterized in that: The function for maximizing customer satisfaction is expressed by minimizing freshness loss and delivery delay: ； In the formula, This represents loss of freshness and delivery delays; This represents the customer's dissatisfaction with the product's freshness. On behalf of clients To capture the maximum loss of freshness in agricultural products; This represents the customer's dissatisfaction with the product delivery time. Representative visits clients The resulting delivery delays; , , representing a collection of customers Represents the number of farms, Represents the total number of customers.

5. The multi-farm harvesting-delivery integrated scheduling method according to claim 1, characterized in that: The dual-objective model for the integrated scheduling of multi-farm harvesting and delivery is expressed as follows: ； ； ； ； ； ； ； ； ； ； ； ； ； ； ； ； ； ； ； ； ； ； ； ； ； ； ； ； ； ； ； ； ； ； ； ； ； In the formula, , representing an agricultural product index, in which , representing a collection of agricultural products, where 0 represents virtual agricultural products. Representative Farm The variety and quantity of agricultural products in China; This represents a set of agricultural products that do not contain 0. , represents the node index, where , representing a set of nodes , representing a collection of farms Represents the number of farms, , representing a collection of customers Represents the total number of customers. Representing the car park, , representing the vehicle index, where , representing a collection of vehicles Represents the number of vehicles; , representing the index of the picking team, in which This represents the gathering of the picking team. Representative Farm The number of picking teams in China; On behalf of clients Total demand; If the customer Farm needed agricultural products but ; otherwise, ; It is a farm agricultural products Freshness loss rate; Represents the vehicle from the node To the node Travel time; This represents the maximum load capacity of each vehicle; On behalf of clients The latest required delivery time; If the farm The picking team After harvesting agricultural products Then came the harvesting of agricultural products. ,but ; otherwise, ; If the farm The picking team After harvesting agricultural products Then came the harvesting of agricultural products. ,but ; otherwise, ; Representative Farm The picking team Picking virtual products Then came the harvesting of agricultural products. ,but ;otherwise, ; If the farm The picking team After harvesting agricultural products Then they picked agricultural products. ,but ; otherwise, ; If the vehicle Serving customers ,but ; otherwise, ; Representative vehicle leave Time; Representative vehicle Arrival at the farm Time; Representative picking farm agricultural products The relative start time; Representative Farm agricultural products The time for completion of harvesting; For farm agricultural products The time for completion of harvesting; If the farm The picking team Picking agricultural products ,but ; otherwise, ; If the farm The picking team Picking agricultural products ,but ; otherwise, ; Representing clients on farms Products in Total demand; Represents infinity; Representative of the picking team For the farm Chinese agricultural products The actual harvesting efficiency; Representative of the picking team For the farm Chinese agricultural products Standard harvesting efficiency; Represents the dynamic harvesting efficiency coefficient; This represents the duration during which the harvesting team maintained peak efficiency without experiencing fatigue; This represents the fatigue duration of the harvesting team; This represents the entire cycle of the harvesting team's efficiency, from peak to decline due to fatigue, and then back to rest and recovery. Representative Farm Chinese agricultural products Delivered to customer The loss of freshness that occurs over time; If the vehicle From the parking lot Heading to the farm ,but Otherwise, ; If the vehicle Access Node Then return to the parking lot ,but Otherwise, ; If the vehicle From the parking lot Head to customer location ,but ; otherwise, ; If the vehicle From the farm Heading to the parking lot ,but ; otherwise, ; If the vehicle Visit customer point Then we went to the farm ,but ;otherwise, ; If the vehicle Access Node Then visit the node ,but ;otherwise, ; Representative vehicle Leaving the parking lot Time; Representative vehicle Leaving the parking lot Time; If the vehicle Serving customers ,but ; otherwise, ; If the vehicle Access Node Then visit the node ,but ; If the vehicle Access Node Then visit the node ,but ; otherwise, ; Indicates if vehicle Visit the farm ,but ; otherwise, ; Representative car field to the farm Travel time; Representative Farm to the farm Travel time; If the vehicle From the farm Go to node ,but Otherwise, ; If the vehicle From the farm Heading to the farm ,but Otherwise, ; Representative vehicle Arrival at the farm Time; Representative vehicle Arrival at the farm Time; Representative vehicles from the farm To the customer point Travel time; Represents the vehicle from the node To the node Travel time; If the vehicle From the farm Head to customer location ,but Otherwise, ; If the vehicle From the customer point Head to customer location ,but Otherwise, ; This indicates that the vehicle has arrived at the customer's location. Time; The vehicle arrived at the customer's location. Time; Representative Team Picking Farm medium-sized products Total harvesting time; Representative Farm Chinese agricultural products Freshness loss rate; On behalf of clients To capture the maximum loss of freshness in agricultural products; Indicates customer The resulting delivery delays.

6. The multi-farm harvesting-delivery integrated scheduling method according to claim 5, characterized in that: The dynamic harvesting efficiency function Represented as: ； In the formula, The time for picking; The duration for which the harvesting team can maintain peak efficiency without experiencing fatigue; This represents the peak efficiency stage for the harvesting team. To reduce the fatigue rate of the harvesting team; The fatigue duration of the harvesting team; This represents the fatigue stage of the harvesting team; The duration of the entire cycle for the harvesting team, from peak efficiency to decreased efficiency due to fatigue, and then to rest and recovery of efficiency. This represents the rest period for the picking team.

7. The multi-farm harvesting-delivery integrated scheduling method according to claim 1, characterized in that: The method for solving the bi-objective optimization model to obtain the optimal solution set includes: Two initial subpopulations are randomly generated: a picking subpopulation and a delivery subpopulation. Each subpopulation contains... An individual represented by an integer encoding; Merging individuals with the same index from two subpopulations yields A complete solution, evaluation Two objective functions are used to obtain a set of optimal solutions as the optimal solution set, which is then stored in an external file. The current harvested seed population and the delivery seed population are input into the trained DQN model, and the DQN model outputs the evolutionary strategy of the current harvested seed population and the delivery seed population. The evolutionary strategy includes three types: crossover and mutation, and crossover and mutation. The two subpopulations after implementing the evolutionary strategy will be used to update the optimal solution set in the external archive. Determine if the maximum running time has been reached. If it has, terminate the process and output the optimal solution set. Otherwise, continue to input the two subpopulations with the updated evolutionary strategy into the trained DQN model until the maximum running time is reached.

8. The multi-farm harvesting-delivery integrated scheduling method according to claim 7, characterized in that: The harvesting subpopulation uses integer encoding, including the task allocation of the harvesting team and the order of agricultural product harvesting; The delivery subpopulation is encoded with integers, including the order of farm and customer visits; The crossover includes uniform crossover, single-point crossover, and sequential crossover. The uniform crossover is only used for the integer corresponding to the task allocation of the picking team in the picking sub-population. The single-point crossover is only used for the integer corresponding to the picking order of agricultural products in the picking sub-population. The sequential crossover is only used for the integer corresponding to the customer index in the delivery sub-population. The mutation is applied only to the integers corresponding to the picking order of agricultural products in the picking subpopulation and the integers corresponding to the customer index in the delivery subpopulation, and the mutation is to swap the positions of the two arbitrary integers.

9. A multi-farm harvesting-delivery integrated scheduling system, characterized in that, include: The configuration module is used to set constraints for the integrated scheduling problem of multi-farm harvesting and delivery, including: each farm grows a variety of agricultural products, and the types of agricultural products grown on each farm are completely different, and they all provide agricultural products to customers; according to customer needs, the harvesting team is arranged to harvest agricultural products in an optimized harvesting order, and vehicles are arranged to depart from the depot to the farm to load agricultural products, deliver agricultural products to customers, and finally return to the depot. Definition module: Used to define two objective functions, including minimizing operating costs and maximizing customer satisfaction. Operating costs include harvesting costs, transportation costs, and vehicle usage costs, while customer satisfaction is related to the freshness of the delivered agricultural products and the delivery time. Establishment Module: Used to establish a dual-objective model for integrated scheduling of multi-farm harvesting and delivery based on the constraints and the objective function; Solution module: Used to solve the bi-objective optimization model to obtain the optimal solution set, and realize the harvesting and distribution of agricultural products based on the harvesting and distribution information in the optimal solution set.

10. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, it implements the steps of the multi-farm harvesting-delivery integrated scheduling method as described in any one of claims 1 to 8.

11. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by the processor, it implements the steps of the multi-farm harvesting-delivery integrated scheduling method as described in any one of claims 1 to 8.

Images