Intelligent scheduling method, system and equipment for wharf production operation tasks and medium

Abstract

The invention belongs to the technical field of intelligent scheduling, and particularly relates to an intelligent scheduling method, system, equipment and medium for wharf production operation tasks, and the method comprises the following steps: dividing an operation task list and corresponding rail-mounted gantry crane information into a plurality of subsets, and distributing each subset to a node, executing genetic algorithm operation on different nodes to enable each node to generate an optimal scheduling scheme; summarizing the optimal scheduling schemes generated by the nodes to form a final scheduling plan; and automatically generating a production operation instruction according to the generated scheduling plan, and distributing the instruction to corresponding rail-mounted gantry crane equipment and container trucks. The distributed computing mode makes full use of computing resources of multiple nodes, and the time for generating the optimal scheduling scheme is greatly shortened. And an optimal scheduling scheme is automatically generated, so that instruction transmission errors or execution deviations caused by human factors are effectively avoided.

Description

Technical Field

[0001] The present invention belongs to the technical field of intelligent scheduling, and particularly relates to an intelligent scheduling method, system, device and medium for terminal production operation tasks. Background Art

[0002] With the rapid growth of container transportation volume, shipping companies have higher and higher requirements for the service level of container terminals, the operation efficiency of logistics and the loading and unloading efficiency of containers. In container terminals, the efficient scheduling of production operation instructions is one of the main factors determining the operation efficiency of terminal equipment and yard operations, and is also related to the terminal service quality such as the in-port time of ships and the on-site time of external container trucks. In the past, terminal production operation instructions mainly relied on the experience of business personnel, combined with the yard situation and equipment operation situation to issue instructions. There are dynamic changes in multiple factors such as different task types, environmental capabilities, and equipment working conditions, and it is impossible to respond in real time, which also leads to unbalanced allocation of yard equipment resources and restricts the exertion of yard machinery resource capabilities.

[0003] In traditional container terminal management, transportation resources such as container trucks (trailers) are often statically configured to fixed production lines, resulting in uneven resource utilization and waste of transportation resources. According to statistics, in the traditional way, the empty running and waiting time exceed 50%, and the utilization rate of equipment resources is not high. In addition, in the operating costs of container terminals, labor costs and equipment usage costs also account for a large proportion, resulting in resource waste. The traditional scheduling methods mainly based on manual scheduling and static scheduling also have problems such as possible errors and safety hazards in manual command and communication, and can no longer meet the needs of the efficient operation of modern ports. Therefore, it is necessary to introduce intelligent scheduling technologies and methods. Summary of the Invention

[0004] Aiming at the problem of uneven allocation of mechanical equipment resources caused by traditional scheduling methods, the present invention provides an intelligent scheduling method, system, device and medium for terminal production operation tasks.

[0005] In a first aspect, the technical solution of the present invention provides an intelligent scheduling method for terminal production operation tasks, including the following steps: Dividing the job task list and the corresponding gantry crane information into several subsets, and allocating each subset to a node, and performing genetic algorithm operations on different nodes to enable each node to generate an optimal scheduling plan; Summarizing and processing the optimal scheduling plans generated by each node to form a final scheduling plan; Automatically generating production operation instructions according to the generated scheduling plan, and dispatching the instructions to the corresponding gantry crane equipment and container trucks.

[0006] As a further limitation of the technical solution of the present invention, the steps for each node to generate an optimal scheduling plan include: Encode the scheduling tasks for each gantry crane and container truck. According to the encoding rules, randomly generate a set number of chromosomes on the node to form an initial population. Determine the fitness function according to the actual requirements of terminal production operations. Calculate the fitness of each individual in the population. Calculate the selection probability of each individual, and select individuals with fitness greater than a set threshold from the population to enter the next generation population according to the selection probability. Randomly select two individuals from the population after the selection operation as parents, and exchange the two parent chromosomes at the crossover point to generate two new offspring chromosomes. Mutate the offspring chromosomes after the crossover operation with a set mutation probability; when the number of iterations of the genetic algorithm reaches the set number, stop the algorithm. At this time, the individual with the highest fitness in the current population is the optimal scheduling plan generated for the node.

[0007] As a further limitation of the technical solution of the present invention, the steps of encoding the scheduling tasks for each gantry crane and container truck include: Adopt an integer encoding method to digitally represent the operation sequence of the gantry crane and the transportation path of the container truck according to set rules. Each chromosome can be represented as an integer sequence of length n, and each integer in the sequence corresponds to the number of a gantry crane. According to the number k of container trucks, adopt the sequential allocation rule to arrange the numbers of the transportation tasks responsible for each container truck in sequence to form a chromosome, and present the sequence and allocation situation of the container truck transportation tasks in a digital form.

[0008] As a further limitation of the technical solution of the present invention, in the step of calculating the selection probability of each individual and selecting individuals with fitness greater than a set threshold from the population according to the selection probability, the formula for calculating the selection probability is as follows: Let the individual in the population have a fitness value of , then its selection probability ; P is the number of chromosomes in the initial population.

[0009] As a further limitation of the technical solution of the present invention, the steps of determining the fitness function according to the actual requirements of terminal production operations include: Taking the maximization of operation efficiency E, the balance of mechanical equipment resource utilization rate U, and the minimization of container truck waiting time W as the goals, determine the fitness function , where is the weight coefficient, and .

[0010] As a further limitation of the technical solution of the present invention, the operation efficiency ; In the formula, n is the number of operation tasks, is the operation duration of the th operation task, is the start time of the th operation task, is the completion time of the th operation task, T is the time period. It should be noted that the value of is the number of operation tasks n; The utilization rate of mechanical equipment resources ; In the formula, m is the number of gantry crane equipment, is the working duration of each gantry crane equipment j within the time period T, is the average working duration; The waiting time of the truck ; In the formula, is the arrival time of each truck l, is the start time of the truck's loading and unloading operation, is the waiting time of the truck.

[0011] As a further limitation of the technical solution of the present invention, the steps of summarizing the optimal scheduling plans generated by each node to form the final scheduling plan include: Collect the optimal scheduling plans generated by each node; According to the actual operation requirements of the terminal, determine the priorities of different evaluation indicators, randomly adjust the task allocation or operation sequence in the plan to generate a new candidate plan, calculate the fitness value of the new plan, and accept the plan with a fitness value less than the set value according to the set probability. After ensuring that there are no conflicts in the task allocation of gantry cranes and trucks according to the rule set information, finally obtain an optimal scheduling plan.

[0012] Present the final scheduling plan in a visual form such as a chart or Gantt chart, providing intuitive and clear operation arrangement information for the terminal management personnel. The management personnel can clearly understand the operation sequence, time arrangement and task allocation of gantry cranes and trucks at a glance, which is convenient for timely discovering potential problems and making adjustments. This visual management method helps to improve the scientificity and timeliness of management decisions and enhance the terminal operation management level.

[0013] Second aspect, the technical solution of the present invention also provides an intelligent scheduling system for terminal production operation tasks, including an intelligent scheduling server and several distributed nodes, each node is communicatively connected to the intelligent scheduling server; each node is provided with an optimal solution generation module; the intelligent scheduling server is provided with a final scheduling plan generation module and an instruction generation and distribution module; The optimal solution generation module is configured to split the job task list and the corresponding gantry crane information into several subsets, and allocate each subset to a node, and perform genetic algorithm operations on different nodes, so that each node generates an optimal scheduling plan; The final scheduling plan generation module is configured to summarize and process the optimal scheduling plans generated by each node to form a final scheduling plan; The instruction generation and distribution module is configured to automatically generate production operation instructions according to the generated scheduling plan, and distribute the instructions to the corresponding gantry crane equipment and container trucks.

[0014] As a further limitation of the technical solution of the present invention, the optimal solution generation module includes an encoding unit, an initialization unit, a fitness function determination unit, a fitness calculation unit, a selection processing unit, a crossover processing unit, a mutation processing unit, and an output unit; The encoding unit is configured to encode the scheduling tasks of each gantry crane and container truck; The initialization unit is configured to randomly generate a set number of chromosomes on the node according to the encoding rule to form an initial population; The fitness function determination unit is configured to determine the fitness function according to the actual requirements of terminal production operations; The fitness calculation unit is configured to calculate the fitness of each individual in the population; The selection processing unit is configured to calculate the selection probability of each individual, and select individuals with fitness greater than a set threshold from the population to enter the next generation population according to the selection probability; The crossover processing unit is configured to randomly select two individuals from the population after the selection operation as parents, and exchange the two parent chromosomes at the crossover point to generate two new offspring chromosomes; The mutation processing unit is configured to mutate the offspring chromosomes after the crossover operation with a set mutation probability; The output unit is configured to stop the algorithm when the number of iterations of the genetic algorithm reaches a set number. At this time, the individual with the highest fitness in the current population is the optimal scheduling plan generated by the node.

[0015] As a further limitation of the technical solution of the present invention, the coding unit is specifically configured to use an integer coding method to digitally represent the operation sequence of the rail-mounted crane and the transportation path of the container truck according to a set rule. Each chromosome can be represented as an integer sequence with a length of n, and each integer in the sequence corresponds to the number of a rail-mounted crane; according to the number k of container trucks, using the sequential allocation rule, the numbers of the transportation tasks responsible for each container truck are arranged in sequence to form a chromosome, and the sequence and allocation situation of the container truck transportation tasks are presented in a digital form.

[0016] As a further limitation of the technical solution of the present invention, the formula for calculating the selection probability is as follows: Let the fitness value of the individual in the population be , then its selection probability ; P is the number of chromosomes in the initial population.

[0017] As a further limitation of the technical solution of the present invention, the fitness function determination unit is configured to determine the fitness function with the maximization of the operation efficiency E, the balance of the utilization rate U of mechanical equipment resources, and the minimization of the waiting time W of the container truck as the goals, where is the weight coefficient, and .

[0018] As a further limitation of the technical solution of the present invention, the operation efficiency ; In the formula, n is the number of operation tasks, is the operation duration, is the start time of the operation, is the completion time of the operation, and T is the time period; The utilization rate of mechanical equipment resources ; In the formula, m is the number of rail-mounted crane equipment, is the working duration of each rail-mounted crane equipment j within the time period T, is the average working duration; The waiting time of the container truck ; In the formula, is the arrival time of each container truck l, is the start time of the container truck's loading and unloading operation, is the waiting time of the container truck.

[0019] As a further limitation of the technical solution of the present invention, the final scheduling plan generation module is used to collect the optimal scheduling plans generated by each node; determine the priorities of different evaluation indicators according to the actual operation requirements of the terminal, generate new candidate plans by randomly adjusting the task allocation or operation sequence in the plan, calculate the fitness values of the new plans, and accept the plans with fitness values less than the set value according to the set probability. After ensuring that there are no conflicts in the task allocation of the rail-mounted gantry cranes and container trucks according to the rule set information, an optimal scheduling plan is finally obtained.

[0020] In a third aspect, the technical solution of the present invention further provides an electronic device, which includes: at least one processor; and a memory communicatively connected to the at least one processor; the memory stores computer program instructions executable by the at least one processor, and the computer program instructions are executed by the at least one processor to enable the at least one processor to execute the intelligent scheduling method for terminal production operation tasks as described in the first aspect.

[0021] In a fourth aspect, the technical solution of the present invention further provides a non-transitory computer-readable storage medium, which stores computer instructions that cause the computer to execute the intelligent scheduling method for terminal production operation tasks as described in the first aspect.

[0022] The beneficial effects of the technical solution of the present invention are as follows: The job task list and the rail-mounted gantry crane information are segmented into subsets and allocated to different nodes, and each node executes the genetic algorithm operation in parallel. This distributed computing method makes full use of the computing resources of multiple nodes and greatly shortens the time for generating the optimal scheduling plan. Compared with the traditional centralized computing and scheduling method, when facing large-scale and complex terminal production operation tasks, it can obtain the scheduling result faster, significantly improve the overall operation efficiency of the terminal, reduce the waiting time for cargo handling, and accelerate the ship turnover speed.

[0023] Automatically generate production operation instructions according to the final scheduling plan and distribute them to the corresponding equipment, realizing seamless docking from the scheduling plan to the actual operation execution. This process reduces manual intervention, effectively avoids instruction transmission errors or execution deviations caused by human factors, improves the accuracy and consistency of operation execution, and further ensures the efficient and orderly progress of terminal production operations. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] In order to more clearly illustrate the technical solution of the present invention, the drawings required to be used in the description will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0025] Figure 1 A schematic flow chart of a method provided in an embodiment of the present invention.

[0026] Figure 2 This is an example of production data used in the embodiment of the present invention.

[0027] Figure 3 This is an example of a yard operation sequence generated by an embodiment of the present invention.

[0028] Figure 4 It is a flow chart of the genetic algorithm in an embodiment of the present invention.

[0029] Figure 5 It is a schematic block diagram of a system provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0030] In order to make the purpose, features and advantages of the present invention more obvious and easy to understand, the technical scheme of the present invention will be clearly and completely described below in conjunction with the drawings in this specific embodiment. Obviously, the embodiments described below are only part of the embodiments of the present invention, not all of them. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0031] like Figure 1 As shown, an embodiment of the present invention provides an intelligent scheduling method for dock production operation tasks, comprising the following steps: S1: Divide the job task list and the corresponding rail crane information into several subsets, assign each subset to a node, and perform genetic algorithm operations on different nodes to generate an optimal scheduling solution for each node; S2: Summarize the optimal scheduling solutions generated by each node to form a final scheduling plan; S3: Automatically generate production operation instructions according to the generated scheduling plan, and distribute the instructions to the corresponding rail crane equipment and container trucks.

[0032] In some embodiments, Figure 4 As shown, the steps for each node to generate the optimal scheduling solution include: S21: Encode the scheduling task for each rail crane and container truck; Integer coding is used to digitally represent the operation sequence of the rail crane and the transportation path of the container truck according to the set rules. Each chromosome can be represented as an integer sequence of length n, and each integer in the sequence corresponds to the number of a rail crane. According to the number k of shuttle carriers, using the sequential allocation rule, arrange the numbers of the transportation tasks responsible for each shuttle carrier in sequence to form a chromosome, presenting the sequence and allocation of the shuttle carrier transportation tasks in a digital form.

[0033] Number the transportation tasks of the shuttle carriers. Assume there are q transportation tasks in total. Then, according to the number of shuttle carriers and the transportation task allocation rule, determine the coding method. For example, if the sequential allocation rule is adopted, the numbers of the transportation tasks responsible for each shuttle carrier can be arranged in sequence to form a chromosome. Assume shuttle carrier 1 is responsible for tasks 1, 3, 5, and shuttle carrier 2 is responsible for tasks 2, 4. Then the encoded chromosome can be expressed as [1, 3, 5, 2, 4]. In this way, through this coding method, the sequence and allocation of the shuttle carrier transportation tasks are presented in a digital form.

[0034] S22: According to the coding rule, randomly generate a set number of chromosomes at the node to form the initial population; S23: Determine the fitness function according to the actual requirements of the terminal production operation; Taking the maximization of the operation efficiency E, the balance of the utilization rate U of mechanical equipment resources, and the minimization of the waiting time W of shuttle carriers as the goals, determine the fitness function , where is the weight coefficient, and .

[0035] Operation efficiency ; In the formula, n is the number of operation tasks, is the operation duration of the th operation item, is the start time of the th operation item, is the completion time of the th operation item, T is the time period. It should be noted that is the number of operation tasks n.

[0036] Utilization rate of mechanical equipment resources ; In the formula, m is the number of gantry crane equipment, is the working duration of each gantry crane equipment j within the time period T, is the average working duration; Waiting time of shuttle carrier ; In the formula, is the arrival time of each shuttle carrier l, is the start time of the shuttle carrier for loading and unloading operations, is the waiting time of the shuttle carrier.

[0037] S24: Calculate the fitness of each individual in the population; S25: Calculate the selection probability of each individual, and select the individuals with fitness greater than the set threshold from the population to enter the next-generation population; The formula for calculating the selection probability is as follows: Let the individual in the population have a fitness value of , then its selection probability ; P is the number of chromosomes in the initial population.

[0038] S26: Randomly select two individuals from the population after the selection operation as parents, and exchange the two parent chromosomes at the crossover point to generate two new offspring chromosomes; S27: Mutate the offspring chromosomes after the crossover operation with a set mutation probability; S28: Whether the termination condition is met; If yes, execute S29; If no, execute S25; S29: The individual with the highest fitness in the current population is the optimal scheduling scheme generated by this node.

[0039] Assume that the job task list contains multiple tasks, and each task has attributes such as task number, job type (such as loading / unloading ships, stacking, etc.), operation time, and cargo weight. The tasks can be segmented according to their attributes. For example, the tasks can be divided into different subsets according to the job type, or divided according to the time window of the operation. For example, the tasks that need to be completed within the next 1 hour are divided into one subset, and the tasks within the next 1 hour are divided into another subset. More complex division methods based on task priority and correlation can also be adopted, and tasks with high priority and mutual correlation are placed in the same subset. The gantry crane information includes gantry crane number, location, working status (idle, busy), lifting capacity, etc. Similarly, the gantry cranes are divided into different subsets according to their locations. For example, the gantry cranes in area A of the terminal are divided into one subset, and the gantry cranes in area B of the terminal are divided into another subset. It is also possible to combine the lifting capacity and working status of the gantry cranes for division, and divide the gantry cranes with similar lifting capacities and currently idle status into the same subset. The segmented job task subsets and the corresponding gantry crane information subsets are put into one-to-one correspondence and assigned to different nodes. These nodes can be different servers or computing units in the cluster. On each node, the genetic algorithm is executed for the assigned job task subset and gantry crane information subset.

[0040] Encode the scheduling plan into chromosomes. For example, an encoding method based on the task order can be adopted. Each gene represents a job task, and the order of the genes represents the sequence of task execution. At the same time, the chromosome also contains information related to the allocation of rail-mounted gantry cranes. For example, assuming there are 5 job tasks in total, the chromosome can be represented as [Task 1, Task 3, Task 2, Task 5, Task 4, Rail-mounted Gantry Crane 1, Rail-mounted Gantry Crane 3, Rail-mounted Gantry Crane 2, Rail-mounted Gantry Crane 1, Rail-mounted Gantry Crane 3], where the first 5 genes represent the task order, and the last 5 genes represent the allocation of the rail-mounted gantry crane corresponding to each task. Randomly generate a certain number of initial chromosomes at each node to form an initial population. The population size can be adjusted according to the scale of the problem and computing resources. For example, it can be set to 100 chromosomes.

[0041] Select better chromosomes from the population according to the fitness value to enter the next generation. The roulette wheel selection method can be used, and the probability of each chromosome being selected is proportional to its fitness value. For example, assume there are 3 chromosomes with fitness values of , and the total fitness value is 1.8. Then the probability of chromosome 1 being selected is 0.8 / 1.8, the probability of chromosome 2 being selected is 0.6 / 1.8, and the probability of chromosome 3 being selected is 0.4 / 1.8.

[0042] Perform crossover operations on the selected chromosomes to generate new chromosomes. For example, the single-point crossover method is adopted. Randomly select a crossover point and exchange part of the genes of the two parent chromosomes at the crossover point to generate two offspring chromosomes. Assume that the parent chromosome 1 is [Task 1, Task 3, Task 2, Task 5, Task 4, Rail-mounted Gantry Crane 1, Rail-mounted Gantry Crane 3, Rail-mounted Gantry Crane 2, Rail-mounted Gantry Crane 1, Rail-mounted Gantry Crane 3], and the parent chromosome 2 is [Task 3, Task 1, Task 5, Task 2, Task 4, Rail-mounted Gantry Crane 3, Rail-mounted Gantry Crane 1, Rail-mounted Gantry Crane 1, Rail-mounted Gantry Crane 2, Rail-mounted Gantry Crane 3]. Randomly select the crossover point as 3. Then the offspring chromosome 1 is [Task 1, Task 3, Task 5, Task 2, Task 4, Rail-mounted Gantry Crane 1, Rail-mounted Gantry Crane 3, Rail-mounted Gantry Crane 1, Rail-mounted Gantry Crane 2, Rail-mounted Gantry Crane 3], and the offspring chromosome 2 is [Task 3, Task 1, Task 2, Task 5, Task 4, Rail-mounted Gantry Crane 3, Rail-mounted Gantry Crane 1, Rail-mounted Gantry Crane 2, Rail-mounted Gantry Crane 1, Rail-mounted Gantry Crane 3].

[0043] Mutate the genes of the chromosome with a certain mutation probability. For example, randomly select a gene for mutation, replace a task with another task or change the allocation of a gantry crane to another gantry crane. Assume the mutation probability is 0.05. If the random number is less than 0.05, mutate a certain gene in the chromosome. For example, mutate task 2 in the chromosome [task 1, task 3, task 2, task 5, task 4, gantry crane 1, gantry crane 3, gantry crane 2, gantry crane 1, gantry crane 3] to task 6, resulting in [task 1, task 3, task 6, task 5, task 4, gantry crane 1, gantry crane 3, gantry crane 2, gantry crane 1, gantry crane 3].

[0044] Repeat the above operations of selection, crossover, and mutation. After a certain number of generations of iteration, the chromosomes in the population are gradually optimized, and finally the optimal scheduling plan for each node is obtained. For example, set the number of iteration generations to 500. When 500 generations are reached, output the chromosome with the highest fitness value in the current population as the optimal scheduling plan for this node.

[0045] In some embodiments, the steps of aggregating the optimal scheduling plans generated by each node to form a final scheduling plan include: collecting the optimal scheduling plans generated by each node; determining the priorities of different evaluation indicators according to the actual operation requirements of the terminal, generating new candidate plans by randomly adjusting the task allocation or operation sequence in the plans, calculating the fitness values of the new plans, and accepting the plans with fitness values less than the set value according to the set probability. After ensuring that there are no conflicts in the task allocation of gantry cranes and yard trucks according to the rule set information, finally obtain an optimal scheduling plan.

[0046] After each node completes the genetic algorithm operation to generate the optimal scheduling plan, send these plans to a summary node. The summary node can be a server dedicated to integrating data. During the aggregation process, check whether there are conflicts between the scheduling plans generated by different nodes. For example, there may be a situation where different nodes allocate the same gantry crane to perform different tasks at the same time, or there is an overlap in the task time arrangements resulting in resource competition.

[0047] Time conflict detection: Check whether there are time conflicts between the scheduling plans generated by different nodes, such as a gantry crane or yard truck being assigned multiple tasks at the same time. For each gantry crane, check its task time arrangements in different plans. If it is found that two tasks both require this gantry crane to execute within the same time period, mark it as a time conflict.

[0048] For time conflicts, make adjustments according to the priorities of the tasks, the impact on operation efficiency, and the utilization rate of equipment resources. Prioritize adjusting the tasks that have less impact on operation efficiency and can make the utilization of equipment resources more balanced.

[0049] Resource conflict detection: Check for resource usage conflicts, including quay cranes, yard trucks, and other possible shared resources. In addition to quay cranes, it is also necessary to check whether there are conflicts in other shared resources (such as yard trucks, etc.). For example, two tasks both need to use the same yard truck to transport goods at the same time.

[0050] For resource conflicts, they can be resolved by increasing resources (such as deploying more yard trucks) or reallocating tasks. For example, reassign one of the tasks that needs to use the conflicting yard truck to another idle yard truck. Consider adjusting the task assignment with the least impact on operation efficiency, or reallocating resources according to the target of equipment resource utilization rate. For example, transfer tasks from equipment with high resource utilization rate to equipment with low utilization rate to achieve balanced resource allocation.

[0051] After conflict detection and resolution, integrate the scheduling plans of all nodes into a final scheduling plan. This scheduling plan should ensure that all job tasks are reasonably arranged in terms of time and resources, avoid conflicts, and at the same time try to meet the various indicators and requirements of terminal production operations. For example, sort the task sequences and quay crane allocation information of each node according to the time sequence and resource allocation situation to form a unified and conflict-free task execution sequence and resource allocation plan as the final scheduling plan.

[0052] Generate specific production operation instructions for each quay crane equipment and yard truck according to the final scheduling plan.

[0053] The content of the quay crane operation instructions includes task number, task type (ship loading / unloading, stacking, etc.), operation start time, operation location (such as ship name, yard location), cargo information (weight, quantity, etc.), operation requirements (such as lifting height, placement location, etc.). For example, for quay crane 1, the generated instruction may be: "Task number 101, ship loading / unloading task, starts at 9:00 am, conducts loading operation at the 3rd hold of ship X, the cargo is steel, with a weight of 50 tons, lifted to stack position 2 in area A of the yard, and placed at a height of 2 meters." The content of the yard truck operation instructions includes task number, task type (transporting goods from the yard to the ship's side or vice versa), transport start time, starting location, destination, transport cargo information, etc. For example, for yard truck 2, the generated instruction may be: "Task number 101, transport task, starts from stack position 2 in area A of the yard at 8:30 am, transports steel to the 3rd hold of ship X, and cooperates with quay crane 1 for loading operation." The generated job instructions are sent to the corresponding rail-mounted gantry cranes and container trucks via a wireless communication network (such as 4G, 5G, Wi-Fi, etc.). Receiving terminals are installed on the rail-mounted gantry cranes and container trucks, which can receive and display these instructions. After the instructions are dispatched, a real-time monitoring mechanism is established to ensure that the rail-mounted gantry cranes and container trucks can receive and execute the instructions in a timely manner. At the same time, the equipment and container trucks can feedback the instruction execution status, such as received, in progress, completed, and other information. If any abnormal situation occurs during the instruction execution process (such as rail-mounted gantry crane failure, container truck congestion, etc.), it is promptly fed back to the dispatching system so that the dispatching system can make adjustments and re-dispatch. For example, if Rail-mounted Gantry Crane 1 fails during the task execution, through the feedback information, the dispatching system can immediately re-arrange another rail-mounted gantry crane to take over the task and adjust the job instructions of the relevant container trucks to ensure the continuity of production operations.

[0054] Furthermore, it should be noted that the information of the rail-mounted gantry crane equipment includes: rail-mounted gantry crane ID, the stacking area where it is located, the maximum load, and the number of rail-mounted gantry cranes in a stacking area.

[0055] Example of rail-mounted gantry crane information: | Crane ID | Stacking Area | Maximum Load | Number of Rail-mounted Gantry Cranes in the Stacking Area | |--------|-------|----------|----------------| | 1| A| 10 tons| 20| | 2| B| 15 tons| 15| |...|...|...|...| | 87| D| 20 tons| 10| Job task list: task ID, starting position, ending position, task type (loading onto ship, unloading from ship, dispatching container, receiving container, moving container), spreader size.

[0056] Example of job task list: | Task ID | Starting Position | Ending Position | Task Type | Spreader Size | |--------|----------|----------|----------|----------| | 1| Stacking Area| Container Truck| Loading onto ship| 40ft| | 2| Container Truck| Stacking Area| Unloading from ship| 20ft| |...|...|...|...|...| Output scheduling plan: The task arrangement for each rail-mounted gantry crane, including task ID, start time, end time, and path.

[0057] The instructions for processing the same quay and the same lane in the rule set include: The priority of the instruction for the container truck to enter the yard first is higher than that of the instruction for loading and unloading the ship.

[0058] The loading instruction can be performed by all container trucks at this quay position.

[0059] The instruction for taking out the container must execute the above instruction first before executing the following instruction.

[0060] The instruction for moving the container between different quays can only be executed by the container trucks in the same quay position and on the same operation lane.

[0061] If it is a double-container instruction, first complete the instruction of one container truck and then perform other instructions.

[0062] If it is a double-container loading instruction, load the front container first and then operate on the rear container.

[0063] The instructions for processing the same quay but different lanes include: The priority of the non-loading / unloading instruction for the container truck to enter the yard first is lower than that of the loading / unloading instruction.

[0064] The instruction for the container truck to reach the quay position has priority for operation.

[0065] For the same type of operation, the one that enters the yard first has priority for operation.

[0066] The priority of different operation types: loading > unloading > sending out the container > receiving the container > moving the container.

[0067] The instructions for processing different quays and different lanes include: The instruction of the current quay position has priority.

[0068] Loading and unloading the ship has priority for operation, followed by receiving and sending the container, and the instruction for moving the container has the lowest priority.

[0069] Loading and unloading the ship gives priority to operating on the one closer to the RMG (Rail Mounted Gantry Crane).

[0070] Receiving and sending the container gives priority to operating on the container truck that enters the yard first.

[0071] The instruction that has already started to be executed should not be cancelled for operation as much as possible.

[0072] If a container truck pulls two containers, give priority to executing the instruction closer to the entrance of the stacking area.

[0073] The priority of the machine to avoid when performing tasks is lower than that of not needing to avoid.

[0074] The yard transfer and priority processing include: If the RMG has no instruction in this stacking area or the container truck has not entered the yard, and there is an instruction in another stacking area, then perform yard transfer.

[0075] The OAS (Operator's Administration System) has the highest priority and must be operated.

[0076] If there is a tipping task for the mission, the tipping task is executed first.

[0077] If there is still a ship loading task on the upper layer of the mission, wait for tipping.

[0078] In the same bay and the same lane, if the driver selects a mission but a gantry truck arrives, the currently selected mission needs to be cancelled.

[0079] The order of selecting missions can be dynamically adjusted.

[0080] Here, the input data includes the birthday data set and feature data; The production data (as Figure 2 shown) includes container information, yard operation records, equipment operation data, etc.; The feature data such as operation time, stack position, equipment type, etc.

[0081] The output is: the optimized yard operation order (as Figure 3 shown), the predicted operation duration, and the operation efficiency improvement report.

[0082] By building an intelligent scheduling model for the container port yard and using machine learning algorithms to train 1 million pieces of production data, the optimization of the yard operation instruction order is achieved. The model realizes the efficient scheduling of yard operations through data preprocessing, feature selection, algorithm training and evaluation. The use of intelligent gantry crane scheduling reduces the complexity and tediousness of manual operations, optimizes the use of equipment and reduces equipment idle time.

[0083] As Figure 5 shown, the embodiment of the present invention also provides an intelligent scheduling system for terminal production operation tasks, including an intelligent scheduling server and several distributed nodes, each node is communicatively connected to the intelligent scheduling server; each node is provided with an optimal solution generation module; the intelligent scheduling server is provided with a final scheduling plan generation module and an instruction generation and distribution module; The optimal solution generation module is used to split the job task list and the corresponding gantry crane information into several subsets, and allocate each subset to a node, and perform genetic algorithm operations on different nodes to make each node generate an optimal scheduling plan; The final scheduling plan generation module is used to summarize and process the optimal scheduling plans generated by each node to form a final scheduling plan; The instruction generation and distribution module is used to automatically generate production operation instructions according to the generated scheduling plan and distribute the instructions to the corresponding gantry crane equipment and gantry trucks.

[0084] In some embodiments, the optimal solution generation module includes an encoding unit, an initialization unit, a fitness function determination unit, a fitness calculation unit, a selection processing unit, a crossover processing unit, a mutation processing unit, and an output unit; The encoding unit is configured to encode the scheduling tasks of each rail-mounted crane and container truck; The initialization unit is configured to randomly generate a set number of chromosomes on the nodes according to the encoding rules to form an initial population; The fitness function determination unit is configured to determine the fitness function according to the actual requirements of the terminal production operation; The fitness calculation unit is configured to calculate the fitness of each individual in the population; The selection processing unit is configured to calculate the selection probability of each individual, and select the individuals with fitness greater than the set threshold from the population to enter the next generation population according to the selection probability; The crossover processing unit is configured to randomly select two individuals from the population after the selection operation as parents, and exchange the two parent chromosomes at the crossover point to generate two new offspring chromosomes; The mutation processing unit is configured to mutate the offspring chromosomes after the crossover operation with a set mutation probability; The output unit is configured to stop the algorithm when the number of iterations of the genetic algorithm reaches the set number. At this time, the individual with the highest fitness in the current population is the optimal scheduling solution generated by the node.

[0085] In some embodiments, the encoding unit is specifically configured to adopt an integer encoding method to digitally represent the operation sequence of the rail-mounted crane and the transportation path of the container truck according to the set rules. Each chromosome can be represented as an integer sequence with a length of n, and each integer in the sequence corresponds to the number of a rail-mounted crane; according to the number k of container trucks, adopting the sequential allocation rule, the numbers of the transportation tasks responsible for each container truck are arranged in sequence to form a chromosome, and the sequence and allocation of the container truck transportation tasks are presented in a digital form.

[0086] The formula for calculating the selection probability is as follows: Let the individual in the population have a fitness value of , then its selection probability ; P is the number of chromosomes in the initial population.

[0087] In some embodiments, the fitness function determination unit is configured to determine the fitness function with the goal of maximizing the operation efficiency E, balancing the utilization rate U of mechanical equipment resources, and minimizing the waiting time W of container trucks , where is the weight coefficient, and .

[0088] Operation efficiency ; Wherein, n is the number of operation tasks, is the operation duration of the th operation task, is the start time of the th operation task, is the completion time of the th operation task, T is the time period. It should be noted that, the value of is the number of operation tasks n; ; Wherein, m is the number of gantry crane equipment, is the working duration of each gantry crane equipment j within the time period T, is the average working duration; Truck waiting time ; Wherein, is the arrival time of each truck l, is the start time of the truck loading and unloading operation, is the truck waiting time.

[0089] Here, the initial weight coefficient is determined according to the long-term experience and management objectives of port operation. For example, if the current port pays more attention to improving operation efficiency to cope with the busy traffic volume, the value of can be appropriately increased, such as setting to 0.5, to 0.3,

[0090] When training the scheduling plan, it is necessary to extract the operation data within a certain period (such as one year) from the port's operation and management data. These data should include the detailed information of each operation task, specifically including the operation type, start time of the operation , end time of the operation , the designed gantry crane equipment number and working duration , arrival time of the truck and waiting time

[0091] Organize the collected data values into a format suitable for input to the training algorithm, use operation efficiency, mechanical equipment resource utilization rate, and truck waiting time as feature variables, and perform training optimization to finally output the scheduling plan.

[0092] In some embodiments, the final scheduling scheme generation module is configured to collect the optimal scheduling schemes generated by each node; determine the priorities of different evaluation metrics according to the actual operation requirements of the terminal, generate new candidate schemes by randomly adjusting the task allocation or operation sequence in the schemes, calculate the fitness values of the new schemes, and accept the schemes with fitness values less than the set value according to the set probability. After ensuring that there are no conflicts in the task allocation of the rail-mounted gantry cranes and the yard trucks according to the rule set information, an optimal scheduling plan is finally obtained.

[0093] An embodiment of the present invention further provides an electronic device, which includes: a processor, a communication interface, a memory, and a communication bus. Among them, the processor, the communication interface, and the memory complete mutual communication through the communication bus. The communication bus can be used for information transmission between the electronic device and the sensor. The processor can call the logic instructions in the memory to execute the following method: S1: Split the job task list and the corresponding rail-mounted gantry crane information into several subsets, and allocate each subset to a node, and perform genetic algorithm operations on different nodes to enable each node to generate an optimal scheduling scheme; S2: Summarize and process the optimal scheduling schemes generated by each node to form a final scheduling plan; S3: Automatically generate production operation instructions according to the generated scheduling plan, and dispatch the instructions to the corresponding rail-mounted gantry crane equipment and yard trucks.

[0094] In addition, when the logic instructions in the above-mentioned memory are implemented in the form of software functional units and sold or used as an independent product, they can be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of this technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present invention. The foregoing storage medium includes: USB flash drives, mobile hard disks, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), magnetic disks, or optical disks and other various media that can store program codes.

[0095] An embodiment of the present invention provides a non-transitory computer-readable storage medium that stores computer instructions, and the computer instructions cause the computer to execute the method provided by the above method embodiment. For example, it includes: S1: dividing the job task list and the corresponding rail crane information into several subsets, and allocating each subset to a node, performing a genetic algorithm operation on different nodes, so that each node generates an optimal scheduling plan; S2: summarizing the optimal scheduling plans generated by each node to form a final scheduling plan; S3: automatically generating production operation instructions according to the generated scheduling plan, and dispatching the instructions to the corresponding rail crane equipment and container trucks.

[0096] It should be understood that the magnitudes of the sequence numbers of the steps in the above embodiments do not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation to the implementation process of the embodiments of the present invention.

[0097] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present invention. Various modifications to these embodiments will be obvious to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to these embodiments shown herein, but will be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. An intelligent scheduling method for dock production tasks, characterized in that: The steps include: The job task list and the corresponding rail crane information are divided into several subsets, and each subset is assigned to a node. The genetic algorithm operation is performed on different nodes so that each node generates an optimal scheduling solution. Summarize the optimal scheduling solutions generated by each node to form the final scheduling plan; Production operation instructions are automatically generated according to the generated scheduling plan, and the instructions are dispatched to the corresponding rail crane equipment and container trucks.

2. The intelligent scheduling method for dock production tasks according to claim 1 is characterized in that: The steps for each node to generate the optimal scheduling solution include: Encode the dispatching tasks for each rail crane and container truck; According to the coding rules, a set number of chromosomes are randomly generated on the nodes to form the initial population; Determine the fitness function based on the actual needs of terminal production operations; Calculate the fitness of each individual in the population; Calculate the selection probability of each individual, and select individuals with fitness greater than the set threshold from the population according to the selection probability to enter the next generation population; Randomly select two individuals from the population after the selection operation as parents, exchange the two parent chromosomes at the crossover point, and generate two new daughter chromosomes; The offspring chromosomes after the crossover operation are mutated with the set mutation probability; when the number of iterations of the genetic algorithm reaches the set number, the algorithm is stopped. At this time, the individual with the highest fitness in the current population is the optimal scheduling solution generated for the node.

3. The intelligent scheduling method for dock production tasks according to claim 2 is characterized in that: The steps to encode the dispatching tasks for each rail crane and container truck include: Integer coding is used to digitally represent the operation sequence of the rail crane and the transportation path of the container truck according to the set rules. Each chromosome can be represented as an integer sequence of length n, and each integer in the sequence corresponds to the number of a rail crane. According to the number of container trucks k, the sequential allocation rule is adopted to arrange the transportation task numbers of each container truck in sequence to form a chromosome, and the order and allocation of the container truck transportation tasks are presented in a digital form.

4. The intelligent scheduling method for dock production tasks according to claim 3 is characterized in that: Calculate the selection probability of each individual, and select individuals with fitness greater than the set threshold from the population to enter the next generation population according to the selection probability. The formula for calculating the selection probability is as follows: Assume that individuals in the population The fitness value is , then the probability of selection ; P is the number of chromosomes in the initial population.

5. The intelligent scheduling method for dock production tasks according to claim 4 is characterized in that: According to the actual needs of terminal production operations, the steps to determine the fitness function include: The fitness function is determined with the goal of maximizing operating efficiency E, balancing mechanical equipment resource utilization U, and minimizing container waiting time W. ,in is the weight coefficient, and .

6. The intelligent scheduling method for dock production tasks according to claim 5 is characterized in that: Operation efficiency ; In the formula, n is the number of job tasks, For the The duration of the task, For the The job start time of the job, For the The completion time of each task, T is the time period; Mechanical equipment resource utilization ; In the formula, m is the number of rail crane equipment, is the working time of each rail crane equipment j in the time period T, is the average working hours; Truck waiting time ; In the formula, is the arrival time of each truck l, The time when the container truck starts loading and unloading operations. Waiting time for the collection card.

7. The intelligent scheduling method for dock production tasks according to claim 6 is characterized in that: The steps of summarizing the optimal scheduling solutions generated by each node to form the final scheduling plan include: Collect the optimal scheduling solutions generated by each node; According to the actual operational needs of the terminal, the priorities of different evaluation indicators are determined. By randomly adjusting the task allocation or operation sequence in the plan, new candidate plans are generated and the fitness values ​​of the new plans are calculated. Plans with fitness values ​​less than the set value are accepted according to the set probability. After ensuring that there is no conflict in the task allocation of rail cranes and container trucks based on the rule set information, an optimal scheduling plan is finally obtained.

8. An intelligent scheduling system for dock production tasks, characterized in that: It includes an intelligent scheduling server and several distributed nodes, each of which is connected to the intelligent scheduling server in communication; each node is provided with an optimal solution generation module; the intelligent scheduling server is provided with a final scheduling solution generation module and an instruction generation and dispatching module; The optimal solution generation module is used to divide the job task list and the corresponding rail crane information into several subsets, and assign each subset to a node, and perform genetic algorithm operations on different nodes to generate an optimal scheduling solution for each node; The final scheduling plan generation module is used to summarize the optimal scheduling plans generated by each node to form a final scheduling plan; The instruction generation and dispatching module is used to automatically generate production operation instructions according to the generated scheduling plan, and dispatch the instructions to the corresponding rail crane equipment and container trucks.

9. An electronic device, characterized in that: The electronic device includes: at least one processor; and a memory communicatively connected to the at least one processor; the memory stores computer program instructions executable by the at least one processor, and the computer program instructions are executed by the at least one processor so that the at least one processor can execute the intelligent scheduling method for terminal production operation tasks as described in any one of claims 1 to 7.

10. A non-transitory computer-readable storage medium, characterized in that: The non-transitory computer-readable storage medium stores computer instructions, and the computer instructions enable the computer to execute the intelligent scheduling method for terminal production operation tasks as described in any one of claims 1 to 7.

Images