Container terminal container truck scheduling optimization method oriented to multi-stage multi-equipment cooperation

Abstract

The invention discloses a container terminal container truck scheduling optimization method oriented to multi-stage multi-equipment collaboration. The method comprises the steps that S1, a multi-stage multi-equipment collaboration container truck scheduling optimization model with the minimum ship completion time and the maximum advance time of a ship as the optimization target is established; s2, designing a dynamic task chain coding method; s3, developing an efficient multi-device collaborative multi-task scheduling schedule generation algorithm for a decoding process; s4, designing selection, crossover and mutation operators of the genetic algorithm; and S5, executing a genetic algorithm, optimizing task allocation, and finally determining an optimal scheduling scheme. According to the method, the technical problem that efficient allocation of quay crane, container trucks and field crane tasks is realized by constructing a container truck scheduling mathematical planning model and combining a genetic algorithm is solved. According to the method, the scheduling scheme is gradually optimized through multiple iterations and algorithms, the optimal solution is finally generated, tasks are reasonably distributed, the operation sequence is optimized, and the equipment waiting time and operation delay are reduced.

Description

Technical Field

[0001] The invention relates to a container terminal truck dispatching optimization method, in particular to a container terminal truck dispatching optimization method oriented to multi-stage and multi-equipment collaboration. Background Art

[0002] In the container terminal and logistics industry, the coordinated scheduling of multiple devices and the improvement of operating efficiency have always been key technical challenges. With the growth of global trade volume and the increase in port loading and unloading needs, the demand for coordinated scheduling between multiple devices such as container trucks, quay cranes, yard cranes, and forklifts has become more urgent. Traditional scheduling methods are difficult to cope with complex equipment coordination, dynamic task allocation, and emergencies, which usually causes equipment resources to be idle or waiting, reducing overall operating efficiency and port throughput capacity.

[0003] The multi-equipment scheduling of container terminals faces highly complex synchronization and coordination challenges: the task progress of multiple container trucks in the same stage needs to be highly coordinated to avoid delays caused by resource sharing; there is a strict operation sequence and priority dependency between container trucks and quay cranes, yard cranes, and forklifts. A slight delay may cause the equipment to be idle or waiting in line, reducing overall efficiency; in addition, the progress of the current task of the container truck directly affects the operation of the subsequent stage, and it is necessary to ensure seamless connection of each task. The scheduling system needs to take into account the reasonable allocation of resources between equipment and the forward-looking arrangement of task processes in a dynamic environment to achieve efficient operation of the terminal.

[0004] Container trucks are highly dependent on quay cranes, yard cranes, and forklifts for task operations. However, existing technologies lack effective multi-device coordination strategies, making it difficult to accurately control resource scheduling and operation sequence between devices. At the same time, existing multi-objective optimization methods mainly focus on equipment utilization and lack detailed management of task sequence and time synchronization. This leads to poor task connection between devices, frequent operation conflicts, and a significant decrease in overall operation fluency and resource utilization. Summary of the invention

[0005] The purpose of the present invention is to solve the above-mentioned deficiencies of the prior art and provide a container terminal truck scheduling optimization method for multi-stage and multi-equipment collaboration. The method constructs a truck scheduling mathematical programming model and combines genetic algorithms to achieve efficient allocation of quay crane, truck and yard crane tasks.

[0006] In order to achieve the above-mentioned object, the present invention is designed to provide a container terminal truck dispatch optimization method for multi-stage and multi-equipment collaboration, which is characterized by comprising the following steps:

[0007] Step S1, establishing a multi-stage multi-equipment coordinated container truck scheduling optimization model with the optimization goal of minimizing the ship completion time and the maximum lead time of the ship;

[0008] Among them, the optimization model function is:

[0009]

[0010] in, , represents the delay penalty weight, represents the urgency weight of the ship, Indicates the current time. Indicates the remaining operation time. Indicates the number of remaining tasks on the key road. represents the efficiency of completing the remaining tasks just before the deadline, Indicates the maximum lead time of all ships;

[0011] The constraint function is:

[0012] (1.1)

[0013] This constraint ensures that each task can only be assigned to one truck;

[0014] (1.2)

[0015] This constraint creates a variable and The connection between When, it means the task Assigned to the collection card , then the card set must be satisfied Completing the task There is a follow-up task On the contrary, if , then it means the task Not by card collection Execution, so there is no set card Executed and Tasks Adjacent tasks;

[0016] (1.3)

[0017] This traffic balance constraint is used to ensure that each task Assigned to the collection card When , there is both a predecessor task and a successor task;

[0018] (1.4)

[0019] This constraint defines the initial task of each truck, ensuring that each truck has a unique task as a starting point;

[0020] (1.5)

[0021] This constraint defines the termination task of each container truck, ensuring that each container truck has no subsequent tasks after the terminal task of the path.

[0022] (1.6)

[0023] This constraint means that the end time of the previous truck transportation task cannot be earlier than the start time of the next task.

[0024] (1.7)

[0025] This constraint means that for a loading task, the time when the task starts working on the yard crane side should not be less than the time when the container truck starts the current task plus the travel time from the operating equipment corresponding to the starting point of the current task to the yard crane corresponding to the current task;

[0026] (1.8)

[0027] This constraint means that for a loading task, the time when the task starts working on the quay crane side should not be less than the time when the task is completed on the yard crane side plus the travel time between the yard crane corresponding to the current task and the quay crane corresponding to the current task;

[0028] (1.9)

[0029] This constraint means that for a loading task, the moment when the container truck completes the task is the moment when the current task is completed on the quay crane side;

[0030] (1.10)

[0031] This constraint means that for the unloading task, the time when the task starts working on the quay crane side should not be less than the time when the container truck starts the current task plus the travel time from the operating equipment corresponding to the starting point of the current task to the quay crane corresponding to the current task;

[0032] (1.11)

[0033] This constraint means that for the unloading task, the time when the task starts working on the yard crane side should not be less than the time when the task is completed on the quay crane side plus the travel time of the container truck from the quay crane corresponding to the current task to the yard crane corresponding to the current task;

[0034] (1.12)

[0035] This constraint means that for the ship unloading task, the moment when the container truck completes the task is the moment when the current task is completed on the yard bridge side;

[0036] (1.13)

[0037] This constraint means that the time when the task completes the operation on the field bridge side is the time when the task starts the operation on the field bridge side plus the operation time of the field bridge to execute a single task;

[0038] (1.14)

[0039] This constraint means that the time when the task is completed on the quay crane side is the time when the task starts working on the quay crane side plus the operation time of the quay crane to execute a single task;

[0040] (1.15)

[0041] This constraint means that when different container trucks arrive at the same yard bridge and start queuing to wait for work, the yard bridge will operate in sequence according to the FCFS strategy based on the arrival time of the container trucks.

[0042] (1.16)

[0043] This constraint means that when different container trucks arrive at the same quay crane and start queuing to wait for operations, the quay cranes will operate in sequence according to the FCFS strategy based on the arrival time of the container trucks.

[0044] (1.17)

[0045] This constraint represents the priority constraint of the task. For any two tasks, if there is a constraint on the order of the tasks, the predecessor task must Post-completion tasks The task operation sequence is mainly to avoid conflicts between the quay crane side and the yard crane side.

[0046] Step S2, designing a dynamic task chain encoding method;

[0047] In the dynamic task chain encoding, the left half of the code represents the number of each task, and the right half corresponds to the number of tasks assigned to each set of cards; Among them, the left half, that is, the task number part: the length of the code is the number of tasks , each code value corresponds to a unique task number, this part represents all tasks and is arranged in a certain order; the right half, that is, the task chain length part: the length of the code is the number of collection cards , each code value represents the number of tasks performed by the corresponding set of cards, The sum of the coded values ​​is equal to the number of tasks ; In the actual matching results, the tasks in the left half are divided into Areas, each area corresponds to a task chain of collection cards;

[0048] Step S3, designing a multi-device multi-task scheduling schedule generation algorithm for the decoding process;

[0049] By initializing the first task stage collection of all set cards , and cyclically select the task with the earliest start time to record and update; after each selection, remove the selected task and check whether there are any remaining tasks to be arranged for the current set of trucks; the algorithm continues to iterate until all task stages are scheduled;

[0050] Step S4, designing selection, crossover and mutation operation operators of the genetic algorithm;

[0051] The operation operators mainly include selection operation, crossover operation, mutation operation and elite retention strategy;

[0052] In the selection operation, the probability of selection is determined by the fitness function value of each chromosome of the parent population. The higher the fitness function value, the greater the probability of the chromosome being selected. By introducing the elite retention strategy, several individuals with the highest fitness in the current population are directly copied to the next generation population in each generation of evolution.

[0053] In the crossover operation, the "two-point crossover" method is used for the task number part. First, two crossover points are randomly selected, and the gene fragments between these two positions of each parent chromosome are retained in their respective daughter chromosomes; then, the daughter chromosomes are filled with genes that are not in the crossover interval from the other parent chromosome in order; the task chain length part is directly inherited from each parent chromosome, that is, daughter chromosome 1 completely inherits the task chain length of parent chromosome 1, and daughter chromosome 2 completely inherits the task chain length of parent chromosome 2;

[0054] In the mutation operation, in the task number part, the mutation operation disrupts the existing order by randomly exchanging the task numbers of two positions, thereby generating a new task arrangement; in the task chain length part, the mutation operation randomly adjusts the number of tasks of a certain set card and reallocates tasks from one set card to another;

[0055] Step S5: Execute a genetic algorithm to optimize task allocation and ultimately determine the optimal scheduling solution;

[0056] First, the genetic algorithm constructs a population by randomly generating an initial scheduling plan. When generating the initial population, the task order and task chain length can be randomly generated: the task numbers are randomly arranged to generate the left half of the code; when generating the right half, the initial value of the number of tasks for each truck is , to ensure that the sum of the number of tasks for all collection trucks is , the initial number of tasks for the last container truck is Then, the fitness is calculated through the multi-device multi-task scheduling schedule generation algorithm to evaluate the advantages and disadvantages of these schemes; in the genetic operation, the elite retention strategy is introduced to ensure that the individuals with the highest fitness in each generation of the population are directly copied to the next generation of the population; the scheduling schemes with high fitness are retained through the selection operation, the crossover operation simulates the information reorganization between different schemes to generate new candidate schemes, and the mutation operation explores more possibilities by introducing random changes; after multiple iterations, the algorithm gradually optimizes the scheduling scheme and finally generates the optimal solution.

[0057] The present invention obtains a container terminal truck scheduling optimization method for multi-stage multi-device collaboration. After multiple iterations, the algorithm gradually optimizes the scheduling plan and finally generates the optimal solution, so as to achieve reasonable task allocation, optimize the operation sequence, reduce equipment waiting time and operation delay. Among them, an efficient multi-task scheduling schedule generation algorithm for multi-device collaboration is designed in the decoding process, which estimates the arrival time of the container truck at the quay crane, yard crane, and stacker, the start time and the end time of the operation, determines the operation sequence of the quay crane, yard crane, and stacker, quickly evaluates the fitness, combines the selection crossover mutation operation of the genetic algorithm, optimizes the operation sequence, generates the optimal solution, and reasonably arranges the start time of each task, so that multiple devices can collaborate efficiently, reduce operation delays and invalid waiting time, improve the punctuality of ship berthing operations, and improve the overall operation efficiency and equipment utilization of the terminal. The global search capability of the genetic algorithm and the characteristics of adapting to complex environments enable it to significantly improve the punctuality of ship berthing operations, equipment collaboration efficiency, and overall operation efficiency of the terminal in the optimization of truck scheduling. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] Figure 1 It is a flowchart of the decoding process of the multi-device multi-task scheduling schedule generation algorithm. DETAILED DESCRIPTION

[0059] The following will be combined with the accompanying drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field belong to the scope of protection of the present invention.

[0060] As an implementation mode of the present invention, a container terminal truck scheduling optimization method for multi-stage multi-device collaboration is provided in this embodiment, which includes the following steps:

[0061] Step S1, establish a multi-stage multi-equipment coordinated container truck scheduling optimization model with the optimization objectives of minimizing the ship completion time and the maximum lead time of the ship. Among them, minimize the ship completion time (delay penalty): ensure that each ship completes the task before the scheduled departure time to prevent delays; minimize the maximum lead time of the ship: reduce the waiting time for ships that complete too early to leave, ensure the synchronization of the operation and the departure time, and complete the operation as close to the expected departure time as possible.

[0062] Among them, the optimization model is expressed as a dual objective function:

[0063]

[0064] in, , represents the delay penalty weight, represents the urgency weight of the ship, Indicates the current time. Indicates the remaining operation time. Indicates the number of remaining tasks on the key road. represents the efficiency of completing the remaining tasks just before the deadline, Indicates the maximum lead time of all ships.

[0065] The constraint function is:

[0066] (1.1)

[0067] This constraint ensures that each task can only be assigned to one truck;

[0068] (1.2)

[0069] This constraint creates a variable and The connection between When, it means the task Assigned to the collection card , then the card set must be satisfied Completing the task There is a follow-up task On the contrary, if , then it means the task Not by card collection Execution, so there is no set card Executed and Tasks Adjacent tasks;

[0070] (1.3)

[0071] This traffic balance constraint is used to ensure that each task Assigned to the collection card When , there is both a predecessor task and a successor task;

[0072] (1.4)

[0073] This constraint defines the initial task of each truck, ensuring that each truck has a unique task as a starting point;

[0074] (1.5)

[0075] This constraint defines the termination task of each container truck, ensuring that each container truck has no subsequent tasks after the terminal task of the path.

[0076] (1.6)

[0077] This constraint means that the end time of the previous truck transportation task cannot be earlier than the start time of the next task.

[0078] (1.7)

[0079] This constraint means that for a loading task, the time when the task starts working on the yard crane side should not be less than the time when the container truck starts the current task plus the travel time from the operating equipment corresponding to the starting point of the current task to the yard crane corresponding to the current task;

[0080] (1.8)

[0081] This constraint means that for a loading task, the time when the task starts working on the quay crane side should not be less than the time when the task is completed on the yard crane side plus the travel time between the yard crane corresponding to the current task and the quay crane corresponding to the current task;

[0082] (1.9)

[0083] This constraint means that for a loading task, the moment when the container truck completes the task is the moment when the current task is completed on the quay crane side;

[0084] (1.10)

[0085] This constraint means that for the unloading task, the time when the task starts working on the quay crane side should not be less than the time when the container truck starts the current task plus the travel time from the operating equipment corresponding to the starting point of the current task to the quay crane corresponding to the current task;

[0086] (1.11)

[0087] This constraint means that for the unloading task, the time when the task starts working on the yard crane side should not be less than the time when the task is completed on the quay crane side plus the travel time of the container truck from the quay crane corresponding to the current task to the yard crane corresponding to the current task;

[0088] (1.12)

[0089] This constraint means that for the ship unloading task, the moment when the container truck completes the task is the moment when the current task is completed on the yard bridge side;

[0090] (1.13)

[0091] This constraint means that the time when the task completes the operation on the field bridge side is the time when the task starts the operation on the field bridge side plus the operation time of the field bridge to execute a single task;

[0092] (1.14)

[0093] This constraint means that the time when the task is completed on the quay crane side is the time when the task starts working on the quay crane side plus the operation time of the quay crane to execute a single task;

[0094] (1.15)

[0095] This constraint means that when different container trucks arrive at the same yard bridge and start queuing to wait for work, the yard bridge will operate in sequence according to the FCFS strategy based on the arrival time of the container trucks.

[0096] (1.16)

[0097] This constraint means that when different container trucks arrive at the same quay crane and start queuing to wait for operations, the quay cranes will operate in sequence according to the FCFS strategy based on the arrival time of the container trucks.

[0098] (1.17)

[0099] This constraint represents the priority constraint of the task. For any two tasks, if there is a constraint on the order of the tasks, the predecessor task must Post-completion tasks The task operation sequence is constrained mainly to avoid conflicts between the quay crane side and the yard crane side.

[0100] Step S2, designing a dynamic task chain encoding method;

[0101] In the dynamic task chain encoding, the left half of the code represents the number of each task, and the right half corresponds to the number of tasks assigned to each set of cards. The left half, i.e. the task number part: the length of the code is the number of tasks , each encoding value corresponds to a unique task number, so the left half The encoding value is unique, which represents all tasks and is arranged in a certain order. The right half, i.e. the length of the task chain: the length of the encoding is the number of cards. , that is, the total number of set cards participating in the task. Each coded value represents the number of tasks performed by the corresponding set card. Therefore, the right half The sum of the coded values ​​is equal to the number of tasks (That is, the total number of tasks assigned to each card set should be equal to the total number of tasks.) In the actual matching results, the tasks in the left half will be divided into Areas, each area corresponds to a task chain of collection cards;

[0102] Step S3, designing a multi-device multi-task scheduling schedule generation algorithm for the decoding process;

[0103] Solve the problem of determining the scheduling schedule between multiple trucks, multiple equipment and multiple tasks by initializing the first task stage set of all trucks , and cyclically select the task with the earliest start time to record and update. After each selection, remove the selected task and check whether there are any remaining tasks to be arranged in the current set of trucks; the algorithm continues to iterate until all task stages are scheduled;

[0104] like Figure 1 As shown, specifically:

[0105] enter:

[0106] • : Number of cards

[0107] • : Number of tasks

[0108] • : Card Collection Task chain length

[0109] • : Card Collection Task queue

[0110] =

[0111] • : Task In stage The transportation process of container trucks is divided into two stages: empty and loaded. The loading and unloading equipment includes quay cranes, yard cranes, and forklifts.

[0112] • :Collection Card Execute the task Arrival equipment Time

[0113] step:

[0114] 1.

[0115] / / Initialization The first stage of all tasks for each truck

[0116] 2. While do

[0117] (a) ( , )

[0118] / / Select the device with the earliest start time Loading and unloading tasks

[0119] (b) Recording equipment and card set

[0120] (c) According to the equipment The operation time of the equipment is updated to update the start time of all subsequent tasks. , and update the start time of the subsequent adjacent tasks of the container truck according to the transportation time of the container truck

[0121] (d) /

[0122] (e) /

[0123] (f) If then

[0124] •

[0125] End While

[0126] Output:

[0127] • :Collection Card In stage By device Loading and unloading tasks Start time of operation

[0128] Step S4, designing selection, crossover and mutation operation operators of the genetic algorithm;

[0129] The operation operators mainly include selection operation, crossover operation, mutation operation and elite retention strategy to improve the convergence efficiency and global optimization ability of the algorithm.

[0130] The selection operation adopts the common roulette operation, that is, the probability of selection is determined according to the fitness function value of each chromosome of the parent population. The higher the fitness function value, the greater the probability of the chromosome being selected, thereby ensuring that high-quality individuals are more likely to be retained. In addition, an elite retention strategy is introduced to directly copy several individuals with the highest fitness in the current population to the next generation of population in each generation of evolution, ensuring that the optimal solution is not destroyed by crossover and mutation operations, further improving the global convergence of the algorithm.

[0131] In the crossover operation, we use the "two-point crossover" method for the task number part. First, we randomly select two crossover points and keep the gene fragments between these two positions of each parent chromosome in their respective daughter chromosomes. Then, the daughter chromosomes sequentially fill in the genes that are not in the crossover interval from the other parent chromosome to ensure the uniqueness of the task number. The task chain length part is directly inherited from each parent chromosome, that is, daughter chromosome 1 completely inherits the task chain length of parent chromosome 1, and daughter chromosome 2 completely inherits the task chain length of parent chromosome 2.

[0132] In the mutation operation, in the task number part, the mutation operation disrupts the existing order by randomly swapping the task numbers of two positions, thereby generating a new task arrangement. In the task chain length part, the mutation operation randomly adjusts the number of tasks of a certain set card, reallocates tasks from one set card to another, and ensures that the total number of tasks remains unchanged.

[0133] Step S5: Execute a genetic algorithm to optimize task allocation and ultimately determine the optimal scheduling solution;

[0134] First, the genetic algorithm constructs a population by randomly generating an initial scheduling plan. When generating the initial population, the task order and task chain length can be randomly generated: the task numbers are randomly arranged to generate the left half of the code. When generating the right half, the initial value of the number of tasks for each truck is , to ensure that the sum of the number of tasks for all collection trucks is , the initial number of tasks for the last container truck is . Then, the fitness is calculated through the multi-device multi-task scheduling schedule generation algorithm to evaluate the pros and cons of these solutions. In the genetic operation, the elite retention strategy is introduced to ensure that several individuals with the highest fitness in each generation of the population are directly copied to the next generation of the population, so as to avoid these high-quality solutions from being destroyed during the crossover or mutation process. The scheduling schemes with high fitness are retained through the selection operation, the crossover operation simulates the information reorganization between different schemes to generate new candidate schemes, and the mutation operation explores more possibilities by introducing random changes, effectively avoiding the algorithm from falling into the local optimum. After many iterations, the algorithm gradually optimizes the scheduling scheme and finally generates the optimal solution, realizing the reasonable allocation of tasks, optimizing the operation sequence, and reducing the equipment waiting time and operation delays. The global search capability and the characteristics of adapting to complex environments of the genetic algorithm enable it to significantly improve the punctuality of ship berthing operations, equipment collaboration efficiency and overall terminal operation efficiency in the optimization of container truck scheduling.

[0135] The present invention is not limited to the above-mentioned optimal implementation mode. Anyone can derive other various forms of products under the inspiration of the present invention. However, no matter what changes are made in the shape or structure, all technical solutions that are the same or similar to those of the present application fall within the protection scope of the present invention.

Claims

1. A container terminal truck dispatch optimization method for multi-stage and multi-equipment collaboration, characterized by: Including the following step: Step S1, establishing a multi-stage multi-equipment coordinated container truck scheduling optimization model with the optimization goal of minimizing the ship completion time and the maximum lead time of the ship; Among them, the optimization model function is: ； in, , represents the delay penalty weight, represents the urgency weight of the ship, Indicates the current time. Indicates the remaining operation time. Indicates the number of remaining tasks on the key road. represents the efficiency of completing the remaining tasks just before the deadline, Indicates the maximum lead time of all ships; The constraint function is: （1.1） （1.2） （1.3） （1.4） （1.5） （1.6） （1.7） （1.8） （1.9） （1.10） （1.11） （1.12） （1.13） （1.14） （1.15） （1.16） （1.17） Step S2, designing a dynamic task chain encoding method; In the dynamic task chain encoding, the left half of the code represents the number of each task, and the right half corresponds to the number of tasks assigned to each set of cards; Among them, the left half, that is, the task number part: the length of the code is the number of tasks , each code value corresponds to a unique task number, this part represents all tasks and is arranged in a certain order; the right half, that is, the task chain length part: the length of the code is the number of collection cards , each code value represents the number of tasks performed by the corresponding set of cards, The sum of the coded values ​​is equal to the number of tasks ; In the actual matching results, the tasks in the left half are divided into Areas, each area corresponds to a task chain of collection cards; Step S3, designing a multi-device multi-task scheduling schedule generation algorithm for the decoding process; By initializing the first task stage collection of all set cards , and cyclically select the task with the earliest start time to record and update; after each selection, remove the selected task and check whether there are any remaining tasks to be arranged for the current set of trucks; the algorithm continues to iterate until all task stages are scheduled; Step S4: Design the selection, crossover and mutation operators of the genetic algorithm: The operation operators mainly include selection operation, crossover operation, mutation operation and elite retention strategy; In the selection operation, the probability of selection is determined by the fitness function value of each chromosome of the parent population. The higher the fitness function value, the greater the probability of the chromosome being selected. By introducing the elite retention strategy, several individuals with the highest fitness in the current population are directly copied to the next generation population in each generation of evolution. In the crossover operation, the "two-point crossover" method is used for the task number part. First, two crossover points are randomly selected, and the gene fragments between these two positions of each parent chromosome are retained in their respective daughter chromosomes; then, the daughter chromosomes are sequentially filled with genes that are not in the crossover interval from the other parent chromosome; the task chain length part is directly inherited from each parent chromosome, that is, daughter chromosome 1 completely inherits the task chain length of parent chromosome 1, and daughter chromosome 2 completely inherits the task chain length of parent chromosome 2; In the mutation operation, in the task number part, the mutation operation disrupts the existing order by randomly exchanging the task numbers of two positions, thereby generating a new task arrangement; in the task chain length part, the mutation operation randomly adjusts the number of tasks of a certain set card and reallocates tasks from one set card to another; Step S5: Execute a genetic algorithm to optimize task allocation and ultimately determine the optimal scheduling solution; First, the genetic algorithm constructs a population by randomly generating an initial scheduling plan. When generating the initial population, the task order and task chain length can be randomly generated: the task numbers are randomly arranged to generate the left half of the code; when generating the right half, the initial value of the number of tasks for each truck is , to ensure that the sum of the number of tasks for all collection trucks is , the initial number of tasks for the last container truck is Then, the fitness is calculated through the multi-device multi-task scheduling schedule generation algorithm to evaluate the advantages and disadvantages of these schemes; in the genetic operation, the elite retention strategy is introduced to ensure that the individuals with the highest fitness in each generation of the population are directly copied to the next generation of the population; the scheduling schemes with high fitness are retained through the selection operation, the crossover operation simulates the information reorganization between different schemes to generate new candidate schemes, and the mutation operation explores more possibilities by introducing random changes; after multiple iterations, the algorithm gradually optimizes the scheduling scheme and finally generates the optimal solution.

Images