Device dispatching method, model training method, electronic device, and storage medium

Abstract

Embodiments of the present disclosure relate to the technical field of warehousing. Disclosed are a device dispatching method, a model training method, an electronic device, and a storage medium. The method comprises: acquiring operation data of a plurality of handling devices and task data of a plurality of tasks to be executed, the plurality of handling devices comprising at least one device to be dispatched; processing the operation data and the task data to obtain feature data of the plurality of handling devices and the plurality of said tasks; using a pre-trained time prediction model to analyze the feature data, and determining task durations for the device to be dispatched to execute said tasks; and on the basis of the task durations for the device to be dispatched to execute said tasks, determining a target task from among the plurality of said tasks, and dispatching the device to be dispatched to execute the target task.

Description

Equipment scheduling methods, model training methods, electronic devices and storage media

[0001] This application claims priority to Chinese Patent Application No. 202411515378.5, filed on October 28, 2024, and Chinese Patent Application No. 202511510879.9, filed on October 21, 2025, the entire contents of which are incorporated herein by reference. Technical Field

[0002] The embodiments disclosed herein relate to the field of warehousing technology, specifically to equipment scheduling methods, model training methods, electronic devices, and storage media. Background Technology

[0003] In modern warehousing operations, handling equipment and workstations are core equipment for improving operational efficiency and reducing labor costs, and are widely used in the storage, retrieval, transfer, and picking of goods.

[0004] Since both the handling process of the handling equipment and the operation process of the workstation are dynamic, during operation, the handling equipment may need to wait or take detours due to problems such as path intersections and congestion. The workstation may enter a waiting state because the handling equipment cannot arrive on time, resulting in the need to improve the overall handling efficiency of the warehouse. Summary of the Invention

[0005] Embodiments of this disclosure provide a device scheduling method, a model training method, an electronic device, and a storage medium.

[0006] According to a first aspect of the embodiments of this disclosure, a device scheduling method is provided, the method comprising: acquiring operating data of a plurality of handling devices and task data of a plurality of tasks to be executed, the plurality of handling devices including at least one device to be scheduled; processing the operating data and the task data to obtain feature data of the plurality of handling devices and the plurality of tasks to be executed; analyzing the feature data using a pre-trained time prediction model to determine the task duration for the device to be scheduled to execute the task to be executed; and determining a target task among the plurality of tasks to be executed based on the task duration for the device to be scheduled to execute the task to be executed, so as to schedule the device to be scheduled to execute the target task.

[0007] According to a second aspect of the embodiments of this disclosure, a model training method is provided, the method comprising: acquiring operational sample data of multiple handling devices and task sample data of multiple handling tasks, the task sample data including the actual task duration of the handling devices performing the corresponding handling tasks; processing the operational sample data and the task sample data to obtain training data of the multiple handling devices and the multiple handling tasks; inputting the training data into an initial time prediction model to obtain a predicted task duration of the handling devices performing the handling tasks; and training a pre-trained time prediction model based on the predicted task duration and the actual task duration.

[0008] According to a third aspect of the embodiments of this disclosure, a device scheduling system is provided, the system comprising: a plurality of handling devices, the plurality of handling devices including at least one device to be scheduled; a control device configured to: acquire operating data of the plurality of handling devices and task data of a plurality of tasks to be executed; process the operating data and the task data to obtain feature data of the plurality of handling devices and the plurality of tasks to be executed; analyze the feature data using a pre-trained time prediction model to determine the task duration for the device to be scheduled to execute the task to be executed; and, based on the task duration for the device to be scheduled to execute the task to be executed, determine a target task among the plurality of tasks to be executed, so as to schedule the device to be scheduled to execute the target task.

[0009] According to a fourth aspect of the embodiments of the present disclosure, an electronic device is provided, comprising: a processor; and a memory for storing executable instructions of the processor; wherein the processor is configured to perform operations of the device scheduling method and / or model training method as described above by executing the executable instructions.

[0010] According to a fifth aspect of the embodiments of the present disclosure, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the steps of the device scheduling method and / or model training method as described above. Attached Figure Description

[0011] Figure 1 shows a schematic diagram of an equipment scheduling system provided by some embodiments of the present disclosure;

[0012] Figure 2 shows a schematic diagram of the layout of an operating area provided by some embodiments of this disclosure;

[0013] Figure 3 shows a flowchart of a device scheduling method provided by some embodiments of the present disclosure;

[0014] Figure 4 shows a sub-flowchart of a device scheduling method provided in some embodiments of this disclosure;

[0015] Figure 5 shows a sub-flowchart of another device scheduling method provided in some embodiments of this disclosure;

[0016] Figure 6 shows a sub-flowchart of another device scheduling method provided in some embodiments of this disclosure;

[0017] Figure 7 shows a sub-flowchart of another device scheduling method provided in some embodiments of this disclosure;

[0018] Figure 8 shows a sub-flowchart of another device scheduling method provided in some embodiments of this disclosure;

[0019] Figure 9 shows a flowchart of a model training method provided by some embodiments of this disclosure;

[0020] Figure 10 shows a sub-flowchart of a model training method provided by some embodiments of this disclosure;

[0021] Figure 11A shows a schematic diagram of a warehousing system provided by some embodiments of the present disclosure;

[0022] Figure 11B illustrates an application scenario provided by some embodiments of this disclosure;

[0023] Figure 12 shows a flowchart of another device scheduling method provided by some embodiments of the present disclosure;

[0024] Figure 13 shows a flowchart of a method for determining arrival and departure times provided by some embodiments of the present disclosure;

[0025] Figure 14 shows a flowchart of a method for constructing a queuing timeline provided by some embodiments of the present disclosure;

[0026] Figure 15 shows a flowchart of a method for determining whether a workstation is in an idle state according to some embodiments of the present disclosure;

[0027] Figure 16 shows a flowchart of a method for supplementing a handling device according to some embodiments of the present disclosure;

[0028] Figure 17 shows a flowchart of yet another device scheduling method provided by some embodiments of the present disclosure;

[0029] Figure 18 illustrates a flowchart of determining supplementary handling equipment according to some embodiments of the present disclosure;

[0030] Figure 19A shows a schematic diagram of constructing a queuing timeline according to some embodiments of the present disclosure;

[0031] Figure 19B shows a schematic diagram of a scheduling candidate shelf provided by some embodiments of the present disclosure;

[0032] Figure 20 shows a schematic diagram of an equipment scheduling device provided in some embodiments of the present disclosure;

[0033] Figure 21 is a schematic diagram of the structure of an electronic device provided in some embodiments of this disclosure. Detailed Implementation

[0034] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein.

[0035] The embodiments of this disclosure provide an equipment scheduling system that can improve the operating efficiency of global handling equipment. The equipment scheduling system provided by the embodiments of this disclosure will be described below with reference to the accompanying drawings.

[0036] Figure 1 is a schematic diagram of an equipment scheduling system provided in some embodiments of this disclosure. As shown in Figure 1, the equipment scheduling system 100 may include a plurality of first handling devices 110 and a control device 120. Among the plurality of first handling devices 110, at least one device is to be scheduled. The device to be scheduled refers to a first handling device among the plurality of first handling devices 110 that has not been assigned a handling task.

[0037] Among the multiple first handling devices 110, each first handling device can be used to perform different handling tasks in the operating area, such as acquiring a container to be handled and moving the container to a designated location. The container to be handled can be a pallet, box, bin, turnover box, special container, etc.

[0038] To illustrate the scheduling of the first handling equipment in this embodiment, Figure 2 shows a schematic diagram of an operating area provided by some embodiments of this disclosure. As shown in Figure 2, the operating area 200 may include a container storage area 210, multiple first handling devices 110, and a workstation 220. The container storage area 210 is used to place containers for storing goods. For example, the container storage area 210 may be equipped with multiple shelves, each shelf having multiple storage locations, and each storage location being used to place one container.

[0039] Workstation 220 includes workstation 1, workstation 2 and workstation 3. Each workstation can be used to pick and sort goods from containers transported by the first handling equipment.

[0040] Multiple first handling devices 110 can operate between container storage area 210 and workstation 220. They can take containers out of container storage area 210 and move them to any workstation in workstation 220, or move containers from any workstation to the shelf in container storage area 210 for storage.

[0041] In some embodiments, the plurality of first handling devices 110 may also operate only in the container storage area 210, such as taking containers off one shelf in the container storage area 210 and moving them to a designated location in the container storage area 210, or moving them to a storage location on another shelf, or moving them to other designated areas.

[0042] It should be noted that the schematic diagram of the operating area shown in Figure 2 is only an illustrative example. Depending on business requirements, the operating area of ​​the first handling equipment can have any different layout and structure, and this embodiment does not make any specific limitations on this.

[0043] The control device 120 is a device or tool for monitoring, operating and adjusting the first handling equipment, etc. It can be used to control and manage the first handling equipment operating in the operating area. For example, the control device 120 can obtain the operating information of the first handling equipment among multiple first handling equipment 110 and control each first handling equipment to move to the corresponding position, thereby taking out the container to be handled at the corresponding position and handling it to another position.

[0044] To improve the operating efficiency of the first handling equipment, in some embodiments, the control device 120 can acquire the operating data of multiple first handling equipment 110 and the task data of multiple tasks to be executed. The multiple first handling equipment 110 includes at least one device to be scheduled. Then, the control device can process the operating data and task data to obtain the feature data corresponding to the multiple first handling equipment 110 and the multiple tasks to be executed. Using a pre-trained time prediction model, the feature data is analyzed to determine the task duration for the device to be scheduled to execute the task. Based on the task duration for the device to be scheduled to execute the task, the target task is determined among the multiple tasks to be executed, so as to schedule the device to be scheduled to execute the target task.

[0045] In some embodiments, the operational data of the plurality of first handling devices may include the scheduling status, location information, travel path, and executed tasks and execution status of each of the plurality of first handling devices. The task data may include the task location, task type, and task content of each of the plurality of tasks to be executed. The scheduling status can be used to indicate whether the first handling device has been assigned a corresponding handling task.

[0046] A pre-trained time prediction model is a machine learning model used to predict the duration of tasks to be executed by scheduled equipment. This model can be a deep learning model, a random forest algorithm, a LightGBM algorithm, or a linear regression algorithm, among others. After training, the time prediction model learns the mapping relationship between features and task durations. Therefore, by analyzing the feature data using the pre-trained time prediction model, the task duration of each task to be executed by the first handling equipment can be obtained, allowing for the allocation of appropriate tasks to the scheduled equipment based on the task duration.

[0047] For example, based on the task duration of each task to be executed by the device to be scheduled, the task with the shortest task duration can be identified as the target task, thereby controlling the device to be scheduled to execute the target task. This method can enhance the adaptability between the first handling device and the tasks to be executed, and improve the execution efficiency of the overall handling tasks.

[0048] In some embodiments, for a task to be performed that involves moving a container to a workstation, the control device 120 can determine the task to be performed whose task duration is closest to the remaining time of the task to be performed by the scheduled device and the remaining time of the current task at each workstation as the target task, so that the time for the workstation to wait for the container to arrive is minimized, or the time for the container to wait for the workstation to process after arriving at the corresponding workstation is minimized, thereby improving the work efficiency of the workstation.

[0049] Therefore, the equipment scheduling system provided by the embodiments of this disclosure can analyze feature data through a pre-trained time prediction model before scheduling the first handling equipment to perform a task, determine the task duration of each task to be performed by the equipment to be scheduled, and select tasks with higher execution efficiency as target tasks according to the task duration. This enables the analysis of the compatibility between the first handling equipment and the tasks to be performed, thereby improving the operating efficiency of the first handling equipment and workstations.

[0050] The embodiments of this disclosure also provide a device scheduling method, which can be executed by the control device 120 shown in FIG1, enabling it to acquire operating data of multiple first handling devices and task data of multiple tasks to be executed. The multiple first handling devices include at least one device to be scheduled. Then, the operating data and task data are processed to obtain feature data corresponding to the multiple first handling devices and the multiple tasks to be executed. Then, a pre-trained time prediction model is used to analyze the feature data to determine the task duration for the device to be scheduled to execute the task to be executed. Based on the task duration for the device to be scheduled to execute the task to be executed, a target task is determined among the multiple tasks to be executed, so as to schedule the device to be scheduled to execute the target task.

[0051] By applying this solution, it is possible to analyze the compatibility between the first handling equipment and the task to be performed, thereby improving the overall operating efficiency of the first handling equipment.

[0052] Figure 3 shows a flowchart of a device scheduling method provided by some embodiments of the present disclosure. As shown in Figure 3, the method may include the following steps 310-340:

[0053] Step 310: Obtain the operating data of multiple first handling devices and the task data of multiple tasks to be executed.

[0054] The multiple first handling devices include at least one device to be scheduled. A device to be scheduled refers to a first handling device in the operating area that has not been assigned a handling task. Multiple first handling devices include all first handling devices in the operating area that are both to be scheduled and already scheduled. A task to be executed refers to a handling task that has been generated but not yet executed. This may include tasks such as acquiring containers, handling containers, placing containers, placing goods into containers handled by first handling devices, and removing goods from containers placed on first handling devices and placing them in a designated location.

[0055] In order to schedule the devices to be scheduled among the multiple first handling devices, the control device can first obtain the operating data of the multiple first handling devices and the task data of multiple tasks to be executed, so as to determine the scheduling status, operating status and position of all first handling devices at the current time, as well as the task content of the tasks to be executed, etc., so as to provide data support for subsequent analysis of the feasibility and efficiency of the devices to be scheduled to execute the tasks.

[0056] For example, the operational data of multiple first handling devices may include device operation data of the first handling devices and task execution data of the first handling devices currently performing tasks. The device operation data may include at least one of the following: scheduling status, travel path, location information, and travel speed of the first handling device. The scheduling status may indicate whether the first handling device has been assigned a handling task. When a first handling device has been assigned a handling task and is currently executing it, the task execution data may include task information and / or task execution status of the task currently being executed by the first handling device. The task execution status may indicate whether the task is being executed, whether it has been completed, etc.

[0057] Task data refers to the task information of a task to be executed, which may include at least one of the following: the starting position, the destination position, the task information, and the task type.

[0058] Among them, task information refers to the detailed content of the task to be executed, and task type can indicate the category of the task to be executed, such as whether it is a task to obtain a container or a task to place a container.

[0059] In some embodiments, the operating data of the first handling equipment mainly reflects the status, performance, and operation trajectory of the equipment. The control device can obtain the operating data of the first handling equipment through a wireless communication link established with the first handling equipment.

[0060] In some embodiments, the task data for the task to be executed is automatically generated by the management system according to business rules or created manually. Therefore, the control device can obtain the task data through the data synchronization interface between the warehouse management system, the production execution system, and other systems.

[0061] Step 320: Process the running data and task data to obtain feature data of multiple first handling devices and multiple tasks to be executed.

[0062] After acquiring operational and task data, in order to improve the accuracy of data analysis, the control device can perform data cleaning, transformation, aggregation, and business logic mapping on the operational and task data, transforming the operational and task data into feature data that can be understood by the model.

[0063] For example, the control device can extract key information strongly correlated with "task duration" from operational data and task data, and transform it into structured features to increase feature dimensions.

[0064] Figure 4 shows a sub-flowchart of a device scheduling method provided in some embodiments of this disclosure. As shown in Figure 4, step 320 can be implemented through steps 410-420:

[0065] Step 410: Based on the running data and task data, generate equipment feature data of multiple first handling devices, path feature data of the devices to be scheduled to reach the destination of each task to be executed, and task feature data of the other handling devices among the multiple first handling devices, excluding at least one of the devices to be scheduled mentioned above, currently executing tasks.

[0066] Among them, equipment feature data refers to the feature data of multiple first handling equipment in the equipment dimension, which may include at least one of the scheduling status, travel path, location information and travel speed of each first handling equipment.

[0067] The path feature data may include the travel path of the scheduled equipment to reach the destination of each task to be executed, as well as at least one of the following: equipment heat statistics and the number of feature points in the travel path.

[0068] For example, the travel path of the scheduled equipment to reach the destination of the task is planned by the control device based on the operating status of the first transport equipment in the operating area at the current time. It can be the shortest path for the scheduled equipment to reach the destination of the task from its current location. The equipment heat statistics in the travel path can include the number of first transport equipment in the travel path. Feature points can include inflection points in the travel path, marker points with special markings, etc., and the number of feature points is the number of inflection points, the number of marker points, etc.

[0069] Task characteristic data refers to the characteristic data of other handling equipment currently performing tasks among multiple first handling equipment. It may include task statistics information of other handling equipment currently performing tasks, such as the statistical number of tasks of various task types.

[0070] Table 1 below shows an example of feature data provided in this embodiment, as shown in Table 1:

[0071] Table 1

[0072] Among them, "cornerPointSize" and "nodeheatCount" are features in the path feature data. Depending on the actual needs, the path feature data can also include other features of the driving path, such as path length and the number of straight road segments in the driving path.

[0073] Step 420: Generate feature data based on device feature data, path feature data, and task feature data.

[0074] After obtaining the equipment feature data, path feature data, and task feature data, these feature data can be integrated. For example, the feature data of each first handling device can be correlated to obtain the feature data of the input time prediction model.

[0075] Through steps 410-420 above, multiple sets of feature data can be generated based on the analysis of operational data and task data, thus increasing the amount and richness of feature data and improving the accuracy of subsequent analysis task duration.

[0076] Step 330: Using a pre-trained time prediction model, analyze the feature data to determine the task duration for the device to be scheduled to execute the task.

[0077] The task duration is the time required for the scheduled equipment to execute the task, predicted using a pre-trained time prediction model. In other words, step 330 utilizes the time prediction model to predict the task duration of different tasks performed by the first handling equipment before scheduling, thus enabling more rational and efficient task scheduling.

[0078] After generating feature data, the feature data can be used as input data and fed into a pre-trained time prediction model to predict the task duration for the device to be scheduled to execute the task.

[0079] For example, when the task to be performed is to obtain a container, the task duration is the time required for the device to be scheduled to travel from its current location to the location of the container and place the container on its own containment structure.

[0080] Step 340: Based on the task duration of the task to be executed by the device to be scheduled, determine the target task among multiple tasks to be executed, and schedule the device to be scheduled to execute the target task.

[0081] After obtaining the task duration of the task to be executed by the device to be scheduled, the control device can determine the target task among multiple tasks to be executed according to the task duration, such as determining the task to be executed with the shortest required task duration as the target task, and thus scheduling the device to be scheduled to execute the target task.

[0082] This method allows for the estimation of the execution time of different tasks by the first handling equipment before scheduling the first handling equipment to perform the tasks to be performed. Then, based on the task duration, the appropriate task to be performed for the first handling equipment is selected, thereby reducing the time spent by each first handling equipment in performing the handling task and improving the operating efficiency of the first handling equipment.

[0083] In some embodiments, multiple tasks to be performed may include multiple retrieval tasks, which are used to retrieve containers to be transported, and the task duration may include a first task duration for the scheduled device to perform the retrieval task.

[0084] Among them, the container to be transported can be a container containing the goods required in the order information, as determined by the control device based on the order information placed by the user, or it can be a container that needs to be exchanged during the sorting process.

[0085] For example, in the operating area shown in Figure 2, assuming there are multiple containers to be moved in the container storage area 210, such as container 1, container 2, and container 3, these three containers are placed on different shelves, and multiple container retrieval tasks may include:

[0086] The tasks are: Retrieve Container 1 (to be transported), Retrieve Container 2 (to be transported), and Retrieve Container 3 (to be transported). Retrieve Container refers to the process of the scheduled equipment moving from its current location to the location of the container and placing the container on the scheduled equipment.

[0087] The task duration includes the first task durations t1, t2, and t3 for the devices to be scheduled to execute box retrieval task 1, box retrieval task 2, and box retrieval task 3, respectively.

[0088] Based on this, in step 340, the control device can determine the retrieval task with the shortest first task duration among multiple retrieval tasks as the target task based on the first task duration of each retrieval task performed by the device to be scheduled, thereby scheduling the device to be scheduled to perform the target task so as to retrieve the container to be transported at the container position corresponding to the target task.

[0089] For example, assuming that among the durations t1, t2 and t3 of the first task mentioned above, t2 is the shortest, then the box retrieval task 2 can be determined as the target task so that the device to be scheduled can execute the box retrieval task that can be completed the fastest among multiple box retrieval tasks.

[0090] Therefore, by using the above method, the time required for the first handling equipment to acquire each container to be handled can be predicted before scheduling the first handling equipment to execute the task to be performed. This allows for the selection of a task to be performed quickly for each first handling equipment, thereby improving the overall operating efficiency of the handling equipment.

[0091] In some embodiments, multiple tasks to be executed may include multiple container delivery tasks, which are used to move containers to be transported on the scheduled device to any workstation. For example, in the above embodiment, after the control device selects a target task for the scheduled device based on a first task duration, the scheduled device executes the target task, that is, goes to the location of the container to be transported indicated by the target task, and retrieves the container. During the process of the scheduled device retrieving the container, the system can generate a container delivery task corresponding to the container, which instructs the scheduled device to continue transporting the container to be transported to the workstation after retrieving it.

[0092] The task duration can include the second task duration for the scheduled equipment to perform the corresponding container delivery task. That is, the time required for the scheduled equipment to move the container to be transported from its location to the workstation.

[0093] In this embodiment, after the device to be dispatched acquires the container to be transported, it needs to transport the container to a workstation so that the goods in the container can be picked or sorted at the workstation. That is, the device to be dispatched needs to complete the container delivery task. The container delivery task can instruct the device to transport the container to be transported to a designated workstation, or it can instruct the device to transport the container to be transported to any workstation.

[0094] Therefore, in step 340, Figure 5 shows a sub-flowchart of another device scheduling method provided by some embodiments of this disclosure. As shown in Figure 5, the control device can execute steps 510-530:

[0095] Step 510: When multiple delivery tasks indicate that the containers to be transported should be moved to the same workstation, obtain the first remaining task execution time for the same workstation.

[0096] The first remaining task execution time refers to the time required for the same workstation to complete all tasks scheduled to that workstation. Among all tasks scheduled to that workstation, each task is executed sequentially according to the scheduling order, and the execution status of the tasks may not be the same. For example, some tasks may be in the process of execution, while others may not have been executed yet.

[0097] When multiple delivery tasks instruct the same workstation to move containers to the same workstation, it means that the same workstation needs to process multiple delivery tasks. However, since a workstation can only process delivery tasks sequentially, the control device can first obtain the first remaining task execution time of the workstation in order to facilitate the determination of the target task.

[0098] Step 520: Based on the second task duration of each scheduled device executing the corresponding box delivery task and the first remaining task execution duration of the same workstation, determine the target task among multiple box delivery tasks.

[0099] Based on the duration of the second task and the remaining execution time of the first task of the workstation, the control device can select the target task from multiple box delivery tasks to ensure that the task rhythm of the scheduled equipment and the workstation matches, that is, after the scheduled equipment completes the box delivery task, the workstation can process the containers to be transported delivered by the scheduled equipment as soon as possible.

[0100] In some embodiments, the control device may determine the first time difference corresponding to each box delivery task based on the second task duration of each scheduled device executing the corresponding box delivery task and the first remaining task execution duration of the same workstation, and determine the box delivery task with the smallest first time difference as the target task.

[0101] The first time difference reflects the difference between the duration of the second task and the execution time of the first remaining task. For the delivery task with the smallest first time difference, when the same workstation has completed all the tasks waiting to be executed, the container to be moved indicated by the delivery task has been or can be delivered to the workstation as quickly as possible, so that the container waiting time for the workstation to process is the shortest, or the workstation waiting time for the container to arrive is the shortest, thereby improving the processing efficiency of the entire warehousing operation.

[0102] Step 530: Schedule the device to be scheduled to perform the target task so that the device to be scheduled can move the container to be transported to the corresponding workstation.

[0103] After the target task is determined, the control device can schedule the equipment to be dispatched to move the container to the corresponding workstation for processing.

[0104] By following steps 510-530 above, we can ensure that the time the workstation waits for the container to arrive is minimized, or that the time the workstation waits for the container to process after it arrives is also minimized. This can improve the workstation's efficiency, reduce container waiting time, and enhance task execution efficiency.

[0105] In some embodiments, multiple tasks to be executed may include multiple retrieval tasks and a corresponding delivery task for each retrieval task. The retrieval task is used to retrieve the container to be transported, and the delivery task is used to transport the container to be transported to the workstation.

[0106] The container retrieval task, corresponding to the container delivery task, refers to the task of moving the container to be transported as indicated by the retrieval task to the workstation. For example, in the operating area shown in Figure 2, assuming that for containers 4 and 5 to be transported, and workstations 1, 2, and 3, multiple tasks to be executed can include multiple retrieval tasks and corresponding container delivery tasks. These multiple retrieval tasks can include: Retrieval Task 1 for retrieving container 4, and Retrieval Task 2 for retrieving container 5.

[0107] Each container retrieval task can be accompanied by a container delivery task, which may include: Container Delivery Task 1, which moves container 4 to workstation 1; Container Delivery Task 2, which moves container 4 to workstation 2; Container Delivery Task 3, which moves container 4 to workstation 3; Container Delivery Task 4, which moves container 5 to workstation 1; Container Delivery Task 5, which moves container 5 to workstation 2; and Container Delivery Task 6, which moves container 5 to workstation 3.

[0108] The task duration includes the third task duration for the equipment to be scheduled to perform the box retrieval task and the fourth task duration for performing the box delivery task.

[0109] For example, in the example above, the task duration may include the third task duration t of the device to be scheduled to perform the box retrieval task 1. 31 The duration of the third task in executing the box retrieval task 2 is t. 32 And the duration t of the fourth task of the equipment to be scheduled to perform the box delivery task 1 41 The duration of the fourth task in the delivery task 2 is t. 42 The duration of the fourth task in the delivery task 3 is t. 43 The duration of the fourth task in the box delivery task 4 is t. 44 The duration of the fourth task in the delivery task 5 is t. 45 The duration of the fourth task in the box delivery task 6 is t. 46 .

[0110] Therefore, Figure 6 shows a sub-flowchart of another device scheduling method provided by some embodiments of this disclosure. As shown in Figure 6, in step 340, the control device can execute steps 610-630:

[0111] Step 610: Based on the third task duration of each box retrieval task executed by the device to be scheduled and the fourth task duration of the corresponding box delivery task, determine the total task duration of each task to be executed by the device to be scheduled.

[0112] The execution of tasks by the equipment to be scheduled refers to the equipment to be scheduled performing the box retrieval task and the corresponding box delivery task.

[0113] The control device can determine the total task duration required for the equipment to be scheduled to execute the tasks based on the third task duration of each box retrieval task and the fourth task duration of the corresponding box delivery task.

[0114] For example, in the above example, based on the duration of the third task and the duration of the fourth task, the total task duration for the device to be scheduled to execute box retrieval task 1 and the corresponding box delivery task 1 can be determined to be t. 31 +t 41 The total task time for executing box retrieval task 1 and the corresponding box delivery task 2 is t. 31 +t 42 The total task time for executing box retrieval task 1 and the corresponding box delivery task 3 is t. 31 +t 43 The total task time for executing box retrieval task 2 and the corresponding box delivery task 4 is t. 32 +t 44 The total task time for executing box retrieval task 2 and the corresponding box delivery task 5 is t. 32 +t 45 The total task time for executing box retrieval task 2 and the corresponding box delivery task 6 is t.32 +t 46 .

[0115] By determining the total task duration, the total time required for the equipment to complete the retrieval and delivery of boxes can be assessed before scheduling the equipment.

[0116] Step 620: Based on the total task duration of each task to be executed by the device to be scheduled and the second remaining task execution duration of the workstation corresponding to each task to be executed, determine the target task among multiple tasks to be executed.

[0117] The target task instructs the scheduled equipment to acquire the container to be transported and move it to the target workstation, which can be any one of multiple workstations. The second remaining task execution time refers to the time required for the workstation to complete all tasks scheduled to it. Among all tasks scheduled to the workstation, each task is executed sequentially according to its assigned order.

[0118] When determining the target task, for example, the control device can select the task corresponding to the workstation with the shortest second remaining task execution time as the target task based on the total task duration of the scheduled device for each task and the second remaining task execution time of the workstation corresponding to each task. This ensures that the workstation completes all tasks as soon as possible and enters an idle state, so that new containers can be processed immediately after they arrive.

[0119] In some embodiments, the delivery task corresponding to each retrieval task may instruct the corresponding container to be transported to the same workstation. In this case, the control device can calculate a second time difference between the total task duration of the scheduled device executing each task and the second remaining task execution duration of the workstation corresponding to each task, and then determine the task with the smallest second time difference as the target task.

[0120] The total task duration for a scheduled device to execute each task refers to the total time required for the scheduled device to complete a single task. For example, the total time required for the scheduled device to acquire a container to be transported and move it to a workstation. The second remaining task execution duration for each task refers to the estimated total time required for the workstation associated with each task to complete all scheduled tasks.

[0121] By calculating the second time difference, the task with the smallest difference among all the tasks to be executed can be selected as the target task to be executed by the scheduled device. This ensures that the time for the device to execute the task is as close as possible to the remaining task time of the workstation, avoiding excessive waiting time for the workstation after the container arrives or for the workstation to wait for the container, thereby improving the overall operation efficiency.

[0122] In some embodiments, the delivery task corresponding to each container retrieval task may also instruct the corresponding container to be transported to any workstation. In this case, Figure 7 shows a sub-flowchart of another equipment scheduling method provided in some embodiments of this disclosure. As shown in Figure 7, the control device can execute steps 710-730:

[0123] Step 710: When each delivery task instructs to move the container to be transported to any workstation, calculate the third time difference between the total task duration of the scheduled equipment executing each task and the second remaining task execution duration of the workstation corresponding to each task.

[0124] In this method, there is no fixed workstation restriction for the container delivery task. The total task duration for the scheduled equipment to execute each task can include the total time required after the scheduled equipment executes the container retrieval task, to move the acquired containers to different workstations, and to deliver the containers to the workstations. The second remaining task execution duration for each task refers to the estimated time required for the workstation associated with each task to complete all tasks scheduled to that workstation.

[0125] For example, suppose there are two delivery tasks for the box retrieval task, which instructs the retrieval of container 6 and the delivery of container 6 to workstation A or workstation B respectively. That is, the delivery task instructs the delivery of container 6 to workstation A or workstation B. The control device can calculate the total task time 1 for the scheduled device to perform each task, that is, to retrieve container 6 and deliver it to workstation A, and the total task time 2 for the scheduled device to retrieve container 6 and deliver it to workstation B.

[0126] By combining the second remaining task execution time T1 of workstation A and the second remaining task execution time T2 of workstation B, the third time difference can be calculated to obtain the difference between the total task duration 1 and T1, and the difference between the total task duration 2 and T2.

[0127] Step 720: If there is only one task with a third time difference less than the preset threshold, then the task with the third time difference less than the preset threshold is identified as the target task.

[0128] The preset threshold can be customized according to actual needs.

[0129] After executing step 710, if there is only one task to be executed with a third time difference less than the preset threshold, the control device can directly determine the task to be executed with a third time difference less than the preset threshold as the target task, so as to schedule the scheduled equipment to prioritize the execution of the task with the shortest waiting time for the workstation to arrive at the container and the shortest waiting time for the container to be transported to be processed by the workstation.

[0130] Step 730: If there are multiple tasks with a third time difference less than a preset threshold, the task with the shortest total task duration among the multiple tasks with a third time difference less than the preset threshold shall be identified as the target task.

[0131] When there are multiple tasks to be executed with a time difference of less than a preset threshold, the control device can prioritize the task with the faster execution speed. That is, the task with the shortest total task time among the multiple tasks to be executed is determined as the target task, which can reduce the overall operation time of the equipment and improve the overall efficiency of warehouse picking.

[0132] Step 630: Schedule the device to be scheduled to perform the target task, so as to retrieve the container to be transported from the container location corresponding to the target task and transport the container to be transported to the corresponding workstation.

[0133] After the target task is determined, the control device schedules the equipment to be scheduled to perform the target task, so as to deliver the container to be transported to the corresponding workstation.

[0134] Through steps 610-630 above, the time required for the scheduled equipment to move the container to be transported to different workstations can be predicted before the scheduled equipment is scheduled to perform the task. This allows for the selection of a task that can be completed quickly for each scheduled equipment, reducing the waiting time for workstations and containers, and improving the overall operating efficiency of the transport equipment.

[0135] In fact, the above scheme reflects two stages of container handling: the stage where the scheduled equipment moves from its location to the location of the container to retrieve it, and the stage where the scheduled equipment moves the container from its location to the workstation. This approach increases the granularity of equipment scheduling analysis, enhances the rationality of equipment scheduling, and thus maximizes equipment operating efficiency.

[0136] In order to analyze the task duration of the device to be scheduled to perform the box retrieval task and the corresponding box delivery task, in some embodiments, the pre-trained time prediction model may include a first time prediction model and a second time prediction model, and the feature data includes first feature data and second feature data.

[0137] The first feature data can be the feature data of the device to be scheduled moving from its current location to the location of the container to be transported and placing the container on the device to be scheduled. The second feature data can be the feature data of the device to be scheduled moving the container to be transported from its current location to the corresponding workstation.

[0138] That is, the first feature data and the second feature data can respectively identify different handling stages of the container to be handled. Taking the stage when the scheduling equipment acquires the container to be handled as an example, Table 2 below shows an example of the first feature data provided in this embodiment, as follows:

[0139] Table 2

[0140] It can be seen that the first feature data is a portion of the feature data mentioned above.

[0141] To improve the accuracy of the model's analysis, in some embodiments, the first feature data may also include feature information such as the location of the device to be scheduled, the location of the container to be transported, and the path taken by the device to the container from its location. Correspondingly, in addition to the feature dimensions listed in Table 2 above, the second feature data may also include feature information such as the location of the container to be transported and the path taken by the device to transport the container from its location to different workstations.

[0142] Therefore, Figure 8 shows a sub-flowchart of another device scheduling method provided by some embodiments of this disclosure. As shown in Figure 8, in step 330, the control device can execute steps 810-820:

[0143] Step 810: Using the first-time prediction model, analyze the first feature data to determine the third task duration for the device to be scheduled to perform the box retrieval task.

[0144] The first-time prediction model can be used to predict the task duration for a scheduled device to acquire different containers. Therefore, the first feature data can be input into the first-time prediction model, and the model can be used to output the third task duration for the scheduled device to perform the container retrieval task.

[0145] Step 820: Using the second time prediction model, analyze the second feature data to determine the fourth task duration for the equipment to be scheduled to perform the box delivery task.

[0146] The second time prediction model can be used to predict the task duration for the scheduled equipment to move different containers to different workstations. The second feature data is input into the second time prediction model, which then outputs the fourth task duration for the scheduled equipment to perform the container delivery task.

[0147] Steps 810-820 allow for the use of time prediction models to analyze the task duration of different stages of tasks performed by the scheduled equipment, thereby improving the accuracy and reliability of duration analysis.

[0148] In this embodiment, the first time prediction model and the second time prediction model can be network models trained using different training datasets. For example, the training dataset for the first time prediction model can be the training data generated by multiple first handling devices moving from their current location to the location of the container within a historical time period, and the training dataset for the second time prediction model can be the training data generated by multiple first handling devices moving from the location of the container to the corresponding workstation within a historical time period.

[0149] Obviously, in the training datasets of the two models, the driving position of the first transport device is different, and the corresponding driving path is also different. Compared with the training dataset of the second time prediction model, in the training dataset of the first time prediction model, since the first transport device is in an unloaded state, the driving path of the first transport device is more flexible and the driving speed can be higher.

[0150] In some embodiments, the control device can perform static analysis on the first feature data to determine the third task duration for each scheduled device to perform the container retrieval task. For example, the third task duration for the scheduled device to reach the location of the container to be transported can be calculated based on the minimum distance between the current location and the target location of the scheduled device and the average speed of the scheduled device in the previous 10 minutes.

[0151] In the above embodiments, the control device can acquire the operating data of multiple first handling devices and the task data of multiple tasks to be executed. The multiple first handling devices include at least one device to be scheduled. Then, the operating data and task data are processed to obtain the feature data corresponding to the multiple first handling devices and the multiple tasks to be executed. Then, a pre-trained time prediction model is used to analyze the feature data to determine the task duration for the device to be scheduled to execute the task. Based on the task duration for the device to be scheduled to execute the task, the target task is determined among the multiple tasks to be executed, so as to schedule the device to be scheduled to execute the target task. This can realize the analysis of the adaptability between the first handling devices and the tasks to be executed, and improve the operating efficiency of the overall handling devices.

[0152] Embodiments of this disclosure also provide a model training method, which can be executed by a control device or by a remote server responsible for model training. Figure 9 shows a flowchart of a model training method provided in some embodiments of this disclosure. As shown in Figure 9, it may include the following steps 910-940:

[0153] Step 910: Obtain operational sample data of multiple first handling devices and task sample data of multiple handling tasks.

[0154] The task sample data includes the actual task duration of the first handling equipment performing the corresponding handling task. The operation sample data and task sample data can be relevant data of the first handling equipment and handling tasks collected within a historical time period.

[0155] The operational sample data of multiple first handling devices may include data related to the status, attributes, and performance of each first handling device during historical operation. For example, the operational sample data may include at least one of the following: scheduling status, travel path, location information, and travel speed of the first handling device. When a first handling device is currently assigned a handling task and is executing it, the operational sample data may also include task information and / or task execution status of the currently executing task.

[0156] Task sample data refers to the task information of a transportation task, which may include at least one of the following: the starting location, the destination location, the task information, the task type, and the task priority.

[0157] Step 920: Process the running sample data and task sample data to obtain training data for multiple first handling devices and multiple handling tasks.

[0158] By processing the running sample data and task sample data, the original sample data can be transformed into directly usable structured training data, generating a sample set corresponding to "feature-label".

[0159] To improve the training effect of the model, in some embodiments, the task sample data may further include the starting position and ending position of the first handling device performing the corresponding handling task. Figure 10 shows a sub-flowchart of a model training method provided in some embodiments of this disclosure. As shown in Figure 10, before executing step 920, the control device or remote server may execute the following steps 1010-1020:

[0160] Step 1010: Based on at least one of the starting position, ending position, and the actual task duration of the first handling equipment performing the corresponding handling task, determine whether the sample data of the first handling equipment performing the corresponding handling task is abnormal data.

[0161] The sample data that is considered abnormal includes any of the following:

[0162] (1) If the distance between the starting point and the ending point is greater than the first distance threshold, the actual task time for the first handling equipment to perform the corresponding handling task is less than the first time threshold. The first distance threshold can be set according to the actual situation of the operating area and the operating status of the first handling equipment. The first time threshold can be the minimum travel time corresponding to the distance corresponding to the first distance threshold.

[0163] For example, for a sample data of the first handling device A, if the distance between the starting position and the ending position in the sample data is greater than the first distance threshold, but the actual task time of the first handling device A performing the corresponding handling task is less than the first time threshold, that is, the starting point and the ending point are far apart, but the task time is very short, it is obviously erroneous data, so the sample data is abnormal data.

[0164] (2) If the distance between the starting point and the ending point is less than the second distance threshold, the actual task duration of the first handling equipment performing the corresponding handling task is greater than the second time threshold. Accordingly, the second distance threshold can also be set according to the actual situation of the operating area and the operating status of the first handling equipment. The second time threshold can be the maximum travel time corresponding to the distance corresponding to the second distance threshold.

[0165] For example, for a sample data of the first handling device B, if the distance between the starting position and the ending position in the sample data is less than the second distance threshold, but the actual task duration of the first handling device A performing the corresponding handling task is greater than the second time threshold, that is, the starting point and the ending point are very close, but the task duration is very long, it is obviously erroneous data, so the sample data is also abnormal data.

[0166] (3) The starting position and the ending position are the same. When the starting position and the ending position are the same, the corresponding sample data is not valid task execution data, so the sample data can also be identified as abnormal data.

[0167] (4) The distance between the starting position and the ending position is less than the third distance threshold. The third distance threshold can be set according to the positional relationship between the regions in the running area.

[0168] For example, suppose in a sample data point the starting position is a location in the first handling equipment docking area and the ending position is the location of a container in the container storage area. Then the distance between the starting position and the ending position should not be less than the distance between the first handling equipment docking area and the container storage area. If the distance between the starting position and the ending position in the sample data is less than the distance between the first handling equipment docking area and the container storage area, then the sample data is obviously abnormal data.

[0169] (5) The actual task duration of the first handling equipment performing the corresponding handling task is less than the third time threshold.

[0170] The third time threshold can be set based on the travel speed of the first handling equipment, the task distance of the handling task, etc., or it can be set by statistically analyzing the actual task duration of the first handling equipment in executing historical handling tasks.

[0171] For example, assuming the task distance of the transport task is S1 (greater than 0) and the maximum travel speed of the first transport device is V1 (greater than 0), then the third time threshold can be set to S1 / V1, meaning the minimum time required for the first transport device to execute the transport task is S1 / V1. When the actual task duration of the transport task in the sample data is less than S1 / V1, the sample data can be identified as abnormal data.

[0172] (6) The actual task duration of the first handling equipment performing the corresponding handling task is greater than the fourth time threshold.

[0173] The fourth time threshold can be set by statistically analyzing the actual task duration of the first handling equipment in executing historical handling tasks. For example, assuming the maximum task duration of the first handling equipment in a historical time period is T, then the fourth time threshold can be set to T+α (α is the adjustment weight, greater than 0). If the actual task duration of the first handling equipment in the sample data for executing the corresponding handling task is greater than T+α, the sample data can be determined to be outlier.

[0174] In this method, by analyzing the starting position, ending position and actual task duration of the first handling equipment during the handling process, it is possible to analyze whether the sample data is usable data and realize the analysis of data validity.

[0175] Step 1020: If the sample data is abnormal, delete the sample data.

[0176] By removing outliers from the task sample data, the reliability of the task sample data can be improved, thereby enhancing the training effect and speed of the model.

[0177] Step 930: Input the training data into the initial time prediction model to obtain the predicted task duration for the first handling equipment to perform the handling task.

[0178] The initial time prediction model can include any one or more of the following: random forest algorithm model, LightGBM algorithm model, and linear regression algorithm model.

[0179] The training data is input into the initial time prediction model, and the predictive power of the initial time prediction model is used to output multiple prediction results, namely the predicted task duration.

[0180] Step 940: Train a pre-trained time prediction model based on the predicted task duration and the actual task duration.

[0181] For example, during training, a first loss value can be calculated using a preset loss function based on the predicted task duration and the actual task duration. If the first loss value is greater than a preset loss threshold, the model parameters of the time prediction model are updated, and training is performed again to obtain a new predicted task duration. The second loss value is then calculated again. If the second loss value is greater than a preset loss threshold, the model parameters are updated again, and training continues until the iteration threshold is reached, or the calculated loss value is less than the preset loss threshold, thus obtaining the pre-trained time prediction model.

[0182] Taking the random forest algorithm model as an example, training data can be input into the initial random forest algorithm model. The model first predicts the task duration of each transportation task based on randomly generated parameters, and then outputs a set of predicted task durations. At this time, the prediction results deviate significantly from the actual task durations. The difference between the predicted task durations and the actual task durations can be quantified by loss functions such as mean squared error and mean absolute error. Then, the model parameters are updated, such as increasing the number of trees, selecting the optimal splitting features and thresholds, etc., to enhance the model's ability to fit the features and task durations, optimize model performance, and continue until training is complete.

[0183] According to the model training method provided in this embodiment, training data can be generated by collecting operating sample data of multiple first handling equipment and task sample data of multiple handling tasks. The initial time prediction model can be trained using the training data to obtain a pre-trained time prediction model, which provides model support for predicting the duration of handling tasks performed by the first handling equipment, thereby improving the rationality of equipment scheduling and the efficiency of equipment operation.

[0184] In some embodiments, the task to be performed may include a queuing task, which instructs some of the multiple first handling devices to move the containers to be handled to the workstation for queuing. In this embodiment, the workstation may be equipped with a queuing channel that can accommodate a certain number of handling devices, each of which queues up and waits for the workstation to perform a picking operation. Due to the presence of queuing handling devices, the workstation is always in the process of task execution and will not be idle for a long time, so the workstation has high operating efficiency.

[0185] Based on this, in some embodiments, the step 340 above, which involves scheduling the device to be scheduled to perform the target task, may further include: scheduling at least one second handling device to transport the container to be handled to the workstation for queuing. The device to be scheduled includes the aforementioned at least one second handling device.

[0186] To illustrate the process of the second handling equipment queuing at the workstation, Figure 11A shows a schematic diagram of a warehousing system provided in some embodiments of this disclosure. This warehousing system can be a subsystem of the equipment scheduling system in the above embodiments, and it can include a first device 101, at least one second handling device 102, and at least one workstation 103. The control device includes the first device 101, which can be an RMS (Robot Management System), etc.

[0187] For example, the second handling device 102 shown in FIG11A can be an intelligent warehousing device in a warehousing system, such as an AGV (Automated Guided Vehicle), a picking and handling device, i.e., a robot. The second handling device is a part of a plurality of first handling devices. The number of second handling devices can be one or more, and this disclosure does not limit this.

[0188] Workstation 103 can be a workstation with picking positions, which can be a single picking position type or a dual picking position type. In some embodiments, the workstation can also be a multi-picking position, such as a four-picking position workstation. This disclosure does not limit this.

[0189] The picking station (workstation) of workstation 103 may include at least one queuing channel, and the at least one queuing channel is provided with an operation point, an exit position and multiple entrance positions. In a workstation with two or more picking stations, the picking stations may be arranged side by side, opposite to each other, or in a circular or fan-shaped layout, and this disclosure does not limit this.

[0190] In some embodiments, as shown in FIG11A, the second handling device may be used to carry the picking target 104.

[0191] In some examples, the picking target 104 may include a shelf that can be moved by a mobile robot. The shelf may be multi-layered, and each layer may have storage locations in four directions (i.e., the four sides of the shelf) or two directions. Goods may be placed directly on the storage locations, or containers may be placed on the storage locations. Each container may contain multiple goods, which may be the same SKU or different SKUs. Goods in the same container may correspond to the same picking order or to different picking orders, which is not limited in this disclosure. Goods may also be placed directly on the storage locations, with each good corresponding to a shelf location and each good corresponding to a picking order.

[0192] In other examples, picking target 104 may also include containers (such as the containers to be transported described above), each container corresponding to one picking order, or multiple picking orders. Containers may be cartons, original boxes, packaging boxes, etc., used to hold goods.

[0193] In the embodiments of this disclosure, the first device 101 and the second handling device 102 in the warehousing system can communicate with each other. The first device 101 can obtain the current status of the second handling device 102, the workstation 103, and the picking target 104. The first device 101 can control or manage the second handling device 102. The first device 101 can assign handling tasks to the second handling device 102. The second handling device 102 can provide feedback to the first device 101 on its real-time location, task execution status, abnormal situations, etc. The first device 101 can assign new handling tasks to the second handling device 102 based on its real-time location and task execution status. This disclosure does not limit the scope of the task.

[0194] Figure 11B illustrates an application scenario provided by some embodiments of this disclosure. As shown in Figure 11B, a first device (not shown in the figure) assigns picking targets and second handling equipment to the workstation based on the orders to be picked. The first device constructs a queuing timeline for the second handling equipment arriving at the workstation based on the system's predicted arrival and departure times of the queued second handling equipment. This allows the first device to replenish handling equipment to the workstation when it is idle, preventing situations where people are waiting for goods at the workstation and avoiding interruptions in the flow of queued second handling equipment, thus achieving efficient operation of the workstation.

[0195] In embodiments of this disclosure, the first device 101 may have the functions of assigning picking targets to workstations, assigning second handling equipment (picking and handling equipment / robots) to picking targets, and constructing a queuing timeline for at least one second handling equipment queuing at workstations. For example, the first device may be deployed with an RMS or Autonomous Mobile Robot (AMR) control system, etc.

[0196] In some examples, the first device can be a terminal or a server. For example, when the first device is a terminal, it can include various personal computers, laptops, smartphones, tablets, and portable wearable devices. When the first device is a server, it can be a standalone server or a server cluster consisting of multiple servers; this disclosure does not limit the scope of the embodiments.

[0197] In embodiments of this disclosure, the picking and handling equipment may be a device for moving shelves, such as picking and handling equipment (e.g., picking robots) in a warehousing system.

[0198] In some embodiments, the picking and handling equipment may be a robot for moving shelves, such as a stealthy robot, like a P robot, which is used to move mobile shelves (also known as portable shelves). For example, the P robot can move mobile shelves to the picking area of ​​the workstation for queuing according to the handling task sent by the server.

[0199] In some embodiments, the picking and handling equipment may be a robot for handling containers, such as an RS robot. The RS robot is used to handle containers in the shelf. For example, the RS robot can move containers in the shelf to the picking area of ​​the workstation to queue according to the handling task sent by the server.

[0200] In some embodiments, different functional modules can be deployed on the picking and handling equipment according to different types of robots, such as a communication module for communication, a power supply module for power supply, a camera module for recognizing road conditions and recognizing QR codes on boxes / shelves / landmarks, and a PNC module for planning and controlling the map.

[0201] In the embodiments of this disclosure, the first device can assign picking targets to workstations according to the picking order and assign a second handling device (i.e., a handling robot) to the picking targets. Then, the second handling device will transport the picking targets, such as containers to be transported, to the corresponding workstations for queuing. The picking targets are transported to the queuing channel of the workstations for queuing. If there is an empty picking position in the workstation, the second handling device located in the first queuing position in the queuing channel will transport the corresponding picking target to the empty picking position so that the staff at the workstation can pick the corresponding goods into the order containers according to the picking order.

[0202] In embodiments of this disclosure, the second handling equipment assigned to the workstation by the first device cannot occupy the highway (channel) when queuing, so as to prevent the queuing rack from affecting the passage of other second handling equipment on the highway (channel).

[0203] Figure 12 is a flowchart of another equipment scheduling method provided in some embodiments of this disclosure, applied to a first device in a warehousing system. The first device can be a control device, or a portion of the control devices within the control device. As shown in Figure 12, the method includes the following steps 1201-1204:

[0204] Step 1201: Obtain information on at least one second handling device that is queued at the workstation.

[0205] In some embodiments, the second handling device is a device that carries the container to be handled and moves it to the workstation. The device to be scheduled includes the second handling device. In other words, each second handling device has a corresponding container to be handled.

[0206] In some embodiments, the information of at least one second handling device includes information about the second handling device assigned to the workstation and the corresponding container to be handled. In other words, obtaining the information of at least one second handling device queued at the workstation may involve obtaining the correspondence between the second handling devices queued at the workstation and the containers to be handled.

[0207] In some embodiments, at least one second handling device in the workstation queue is assigned to the workstation by the first device based on the picking order, and each picking order corresponds to at least one picking target, such as a container to be transported.

[0208] For example, the RMS system dispatches robots to move shelves or containers to corresponding workstations for queuing.

[0209] In some embodiments, the first device assigns at least one picking target to the workstation based on multiple pending orders, cargo information of multiple picking targets, and a queuing number threshold of the workstation.

[0210] In some embodiments, at least one picking target corresponds to at least one first order among a plurality of orders to be picked.

[0211] In some embodiments, the queue number threshold for a workstation is determined based on the principle that the queue position of the workstation does not exceed the aisle position, wherein the aisle is a path for the second handling equipment to pass through the handling rack, and the queue number threshold may be the maximum value that the distance between the workstation and the aisle can accommodate the queue of the second handling equipment.

[0212] For example, the queue number threshold is the maximum number of actual queues at the workstation. The queue number threshold needs to ensure that the furthest position of the queuing robot does not block the path (highway) for the robot to pass through the site.

[0213] In some embodiments, the order to be picked can be an order that needs to be picked, which includes the name and quantity of the goods to be picked, and each order to be picked includes at least one name and quantity of the goods to be picked.

[0214] In some embodiments, the goods information of the picking target may be the name, quantity, and corresponding container location of the goods stored on the shelf, and the second handling device may be a device for moving the shelf. Each container may contain the same type of goods or different types of goods. One container may correspond to one picking order or multiple picking orders, which is not limited in this disclosure.

[0215] In some embodiments, the goods information for the picking target can be the name of the goods stored on the shelf and the corresponding shelf location. Each item corresponds to a location on the shelf, and each item corresponds to a picking order.

[0216] In some embodiments, the goods information of the picking target may be the name and quantity of goods in the container, and the second handling device may be a device for handling the container.

[0217] For example, if the queue size threshold is 14, the first device allocates second handling equipment to each workstation based on the current number of workstations and pending orders, ensuring that no more than 14 second handling devices are allocated to each workstation. For instance, in the case of only one workstation, the first device allocates 14 second handling devices to the workstation. Each second handling device is used to handle one shelf, and each shelf corresponds to one pending order, resulting in 14 pending orders. The goods stored on each shelf can also correspond to multiple pending orders, and the same pending order can correspond to goods on multiple queued shelves. Therefore, the number of pending orders corresponding to the shelves handled by the second handling devices allocated to the workstation can be the same as or different from the number of second handling devices.

[0218] For example, in the case of multiple workstations, the first device assigns a second handling device to each workstation according to the order to be picked, so that each workstation has a second handling device, and the number of second handling devices assigned does not exceed the queue number threshold.

[0219] In some embodiments, the first device can determine at least one picking target corresponding to the first order from among multiple picking targets based on the name and quantity of the goods to be picked in multiple picking orders, assign the at least one picking target to a workstation, and assign a second handling device to each picking target as at least one second handling device queuing at the workstation.

[0220] For example, the RMS system assigns orders to workstations and robots based on the order details, the current workstation status, and the location of robots, shelves, or containers. The robots then transfer the corresponding shelves or containers to the workstations.

[0221] Step 202: Based on information from at least one second handling device, determine the arrival time of the at least one second handling device at the workstation and the departure time from the workstation.

[0222] In some embodiments, in step 1202, the estimated arrival time and estimated departure time of at least one second handling device at the workstation can be determined based on at least one second handling device queuing at the workstation and the corresponding picking target, such as a container to be handled. In embodiments of this disclosure, the arrival time and departure time are both estimated by the first device as the time when the second handling device arrives at the workstation and the time when it leaves the workstation.

[0223] In some embodiments, in step 1202, the arrival time and departure time of at least one second transport device can be determined by the status and location information of the second transport device obtained by the first device.

[0224] In some embodiments, in step 1202, the arrival time and departure time of the second handling device can also be determined by obtaining feedback from the second handling device, wherein the second handling device can independently determine its arrival time and departure time from the workstation to inform the first device.

[0225] In some embodiments, in step 1202, the arrival and departure times of the second handling device can also be calculated by using the status and position of the second handling device obtained by the first device and the status and position of the picking target carried by the second handling device.

[0226] In some embodiments, the arrival and departure times determined based on information from at least one second transport device may be the arrival and departure times of the second transport device as predicted by the first device, and the method of determination is not limited in this disclosure.

[0227] Step 1203: Determine whether the workstation is in an idle state based on the arrival and departure times of at least one second handling device.

[0228] In some embodiments, whether a workstation is idle can be determined by whether the second handling equipment queuing at the workstation can keep the workstation staff continuously engaged in picking operations. In other words, the workstation is busy if the queuing second handling equipment can meet the workstation's continuous picking needs; otherwise, the workstation is idle. By ensuring the workstation's continuous picking needs are met, the arrival of second handling equipment at the workstation can be uninterrupted, meaning there will never be a situation where there is no second handling equipment available. This allows the staff at the workstation to perform picking operations continuously. Alternatively, the number of second handling equipment at the workstation at any given time can be such that the staff at the workstation are always engaged in picking operations.

[0229] In some embodiments, the time period during which the second transport device arrives at the workstation can be determined based on the arrival and departure times of at least one second transport device. Then, based on the time period during which each second transport device stays at the workstation, it can be determined whether the workstation is in an idle state. In other words, the time period during which each second transport device stays at the workstation can be determined based on the arrival and departure times of at least one second transport device. Based on the time period during which the device stays, it can be determined whether the workstation is in an idle state.

[0230] In some embodiments, based on the arrival and departure times of at least one second handling device, it can be determined whether other second handling devices have arrived at the workstation and are queuing between the arrival and departure times of any one of the second handling devices, thereby determining whether the workstation is in an idle state.

[0231] In some embodiments, based on the arrival and departure times of at least one second handling device, it can be determined whether the workstation is in an idle state by determining the number of second handling devices queuing at the workstation at the same time and comparing the number of second handling devices at the workstation at the same time with a preset judgment threshold.

[0232] The judgment threshold can be determined by the system based on the number of picking positions and queuing method of the workstation. For example, the judgment threshold can be set to 2. For a workstation with a single picking position and a single queuing channel, when there are two second handling devices on the workstation at the same time, one second handling device is picked at the picking position and the other is waiting in the queuing channel. In this case, the workstation is in a non-idle state. For a workstation with two picking positions and two queuing channels or two picking positions and a single queuing channel, when there are two second handling devices on the workstation at the same time, the two second handling devices are respectively located at the picking positions waiting to be picked. In this case, the workstation is in a non-idle state. Therefore, when the number of second handling devices on the workstation at the same time is less than 2, the workstation is determined to be in an idle state.

[0233] In some embodiments, in determining whether a workstation is idle, the workstation can be determined to be idle if at least one second handling device's time period (the time between arrival and departure times) does not overlap with that of other second handling devices. In other words, when the expected time periods for each second handling device assigned to the workstation do not overlap with each other—that is, the next second handling device arrives at the workstation only after the previous one leaves—the workstation has no second handling devices waiting in the queue, and this state can be determined as idle, meaning the workstation is not operating at full capacity.

[0234] In the above embodiments, the method for determining whether the workstation is in an idle state based on the arrival and departure times of at least one second handling device is not limited in this disclosure.

[0235] Step 1204: When the workstation is idle, replenish the moving equipment to the workstation according to the idle time.

[0236] In some embodiments, when the workstation is idle, the number of second handling devices in the queue cannot fully meet the picking needs of the workstation. For example, after the previous second handling device leaves, it is necessary to wait for a period of time before the next second handling device arrives, during which the workstation is idle.

[0237] In some embodiments, the idle duration can be the length of the time during which the workstation is in an idle state, and the idle duration is equal to the length of the idle time period.

[0238] In some embodiments, if there is an idle period between the arrival and departure times of at least one second handling device, it indicates that the workstation is idle during the idle period.

[0239] In some embodiments, there can be multiple idle time periods. During each idle time period, the first device can supplement the workstation with handling equipment to avoid interruption of the second handling equipment queued by the workstation.

[0240] In some embodiments, an idle state indicates that there are insufficient shelves at the workstation to meet the workstation's picking needs, resulting in a situation where workers are waiting for goods, causing a break in the workstation's picking shelves. For example, if the workstation is a single picking station type, and there is an idle period without a second handling equipment queuing, the workstation will have no second handling equipment queuing during the idle period, leading to workers waiting for goods; or, if the workstation is a dual picking station type, and there is an idle period with only one second handling equipment queuing, it means that the workstation will have one picking station without any shelves or boxes to be picked, resulting in a situation where workers are waiting for goods.

[0241] For example, for a workstation with a dual picking location, a dual picking location can be understood as two robots simultaneously moving the same shelf or box of goods to be picked to the workstation. Then, the staff can take the goods from the two shelves or boxes according to the order to be picked, that is, the staff is working at full capacity. However, if only one shelf or box is moved to the workstation, then after the goods in the current shelf or box are picked, the staff needs to wait for the next second handling equipment to arrive before they can start picking, thus causing the workstation to have a situation where the staff is waiting for the goods.

[0242] In some embodiments, supplementing a workstation with a handling device based on idle time can be achieved by determining a second handling device that can arrive at the workstation during the idle period from a pool of candidate handling devices and candidate picking targets.

[0243] In some embodiments, supplementing a workstation with a handling device can be based on supplementation logic to determine a second handling device from candidate handling devices and candidate picking targets. The supplementation logic can be based on the principle of arrival time priority, that is, the arrival time of the second handling device must be before the start time of the workstation's idle time period, so as to avoid affecting the normal picking order and dwell time period of the second handling device in the queue at the workstation.

[0244] In some embodiments, the supplementary logic for replenishing the workstation with handling equipment can be based on the principle of departure time priority, that is, the departure time of the second handling equipment must be before the arrival time of the next queued second handling equipment, so as to avoid affecting the normal picking order and dwell time of the second handling equipment in the queue at the workstation.

[0245] In some embodiments, the supplementary logic for adding handling equipment to the workstation may be based on the principle of maximizing the number of second handling devices that can arrive at the workstation during the idle period. That is, during the idle period, priority is given to ensuring that the number of second handling devices that can arrive is maximized, so as to ensure that the workstation can pick the maximum amount of goods.

[0246] In some embodiments, when a workstation is idle, supplementing the workstation with handling equipment may be done when there is only one second handling equipment in the queue during the workstation's idle period, and / or when there is no second handling equipment in the queue.

[0247] In some embodiments, supplementary handling equipment may be determined from at least one candidate handling equipment.

[0248] In some embodiments, supplementing the workstation with a handling device may be based on the idle time, by determining a second handling device that can reach the workstation within the idle time from at least one candidate handling device as a supplementary second handling device.

[0249] In the above embodiments, by determining the arrival and departure times of at least one second handling device in the queue at the workstation, it is possible to determine whether the workstation is in an idle state based on the different judgment logics described above. If the workstation is determined to be in an idle state, the second handling device in the queue can be replenished to the workstation in a timely manner to avoid the interruption of the flow of second handling devices arriving at the workstation and causing a situation where people are waiting for goods. At the same time, the operating efficiency of the workstation is further improved.

[0250] Figure 13 is a flowchart of a method for determining arrival and departure times according to some embodiments of this disclosure. Based on the embodiment shown in Figure 12, Figure 13 further describes step 1202 in Figure 12, as shown in Figure 13, including the following steps 1301-1303:

[0251] Step 1301: Obtain location information, average operating speed of the second handling equipment, and estimated picking time.

[0252] In some embodiments, the location information includes the location of at least one second handling device and / or the location of at least one container to be handled.

[0253] In some embodiments, the second handling device has a corresponding container to be handled. The second handling device may have already acquired the container to be handled, or it may not have acquired the container to be handled yet. In other words, the second handling device has identified the container to be handled and is already carrying the corresponding container to be handled, and needs to move the container to be handled to the workstation; or the second handling device has not yet acquired the corresponding container to be handled and needs to move from its current location to the location of the container to be handled in order to move the container to be handled to the workstation.

[0254] For example, the location of the second handling equipment can be its current location before the corresponding shelf has been moved. In this case, the location information to be obtained includes the location of the second handling equipment and the location of the shelf. Alternatively, the location of the second handling equipment can be its current location after the corresponding shelf has been moved. In this case, the location of the second handling equipment and the location of the corresponding shelf are the same, and only the location of the second handling equipment or the location of the shelf needs to be obtained.

[0255] In some embodiments, the average operating speed of the second handling equipment may be the average operating speed of the second handling equipment calculated by the first equipment based on historical data.

[0256] For example, it could be the robot's average running speed.

[0257] In some embodiments, the estimated picking time may be the picking time estimated by the first device based on historical picking data. The estimated picking time may be the average picking time of a shelf or the average picking time of a container. The estimated picking time may be updated periodically. In other words, the first device may calculate the average picking time based on picking data in different time periods and update the estimated picking time periodically.

[0258] Step 1302: Based on location information, average operating speed, and estimated picking time, construct a queuing timeline for at least one second handling device at the workstation, and determine the arrival time of at least one second handling device at the workstation and the departure time of the device from the workstation.

[0259] In some embodiments, the queuing timeline includes the sorting value, arrival time, departure time, and dwell time of at least one second handling device arriving at the workstation.

[0260] In some embodiments, the workstation is located in a fixed position. After the workstation is set up, its location information can be input into the system of the first device for the first device to use. The first device can use the acquired location information, average operating speed, and estimated picking time to build a queuing timeline for at least one second handling device.

[0261] In some embodiments, a queuing timeline is constructed for at least one second handling device at the workstation based on location information, average operating speed, and estimated picking time. This can be done by determining the queuing timeline of at least one second handling device according to the order in which the at least one second handling device arrives at the workstation.

[0262] In some embodiments, a queuing timeline is constructed for at least one second handling device at the workstation based on location information, average operating speed, and estimated picking time. This can be achieved by determining the queuing timeline of at least one second handling device according to the order in which the devices leave the workstation.

[0263] In some embodiments, a queuing timeline is constructed for at least one second handling device at the workstation based on location information, average operating speed, and estimated picking time. This can be achieved by arranging at least one second handling device and determining the order in which the second handling devices arrive at the workstation and wait in line.

[0264] In the above embodiments, the first device constructs a queuing timeline for the workstation by acquiring location information, average operating speed, and estimated picking time, thereby determining the sequence of at least one second handling device and the arrival and departure times of each second handling device.

[0265] Figure 14 is a flowchart of a method for constructing a queuing timeline provided by some embodiments of this disclosure. Based on the embodiments shown in Figures 12 and 13, Figure 14 further describes step 1302 in Figure 13, as shown in Figure 14, which includes the following steps 1401-1403:

[0266] Step 1401: Based on the location information and average operating speed of each of the at least one second handling devices, determine the time when each second handling device arrives at the workstation.

[0267] In some embodiments, when the location of the second handling device is the same as the location of the container to be handled, the arrival time of the second handling device at the workstation is determined based on the location of the container to be handled and the average operating speed of the second handling device.

[0268] In some embodiments, when the location of the second handling equipment is the same as the location of the corresponding shelf or container, the arrival time of the second handling equipment at the workstation is determined based on the location of the shelf or container and the average operating speed of the second handling equipment.

[0269] For example, once the robot has retrieved the shelf or container, the RMS system calculates the time it takes for the shelf or container to reach the workstation based on the location of the shelf or container and the robot's average running speed.

[0270] In some embodiments, when the positions of the second handling device and the container to be handled are different, the arrival time of the second handling device at the workstation is determined based on the position of the second handling device, the position of the container to be handled, and the average operating speed of the second handling device.

[0271] In some embodiments, if the location of the second handling equipment is different from the location of the corresponding shelf or container, the time for the shelf or container to arrive at the workstation is determined based on the location of the second handling equipment, the location of the shelf or container, and the average operating speed of the second handling equipment.

[0272] For example, if the robot has not yet moved the shelf or container, the RMS system can determine the time it will take for the robot to arrive at the workstation after moving the shelf or container, based on the distance from the robot to the shelf or container, the distance from the shelf or container to the workstation, and the robot's average running speed.

[0273] Step 1402: Based on the arrival time at the workstation and the estimated picking time, determine the departure time of each second handling device from the workstation.

[0274] In some embodiments, the time for each second handling device to leave the workstation can be calculated based on the time each second handling device arrives at the workstation and the estimated picking time.

[0275] In some embodiments, after obtaining the arrival time of the second handling device and the container to be handled at the same location in step 1401, the departure time of each second handling device from the workstation can be determined based on the arrival time at the workstation and the estimated picking time.

[0276] In some embodiments, after obtaining the arrival time of the second handling device and the container to be handled at different locations through step 1401, the departure time of each second handling device from the workstation can also be determined based on the arrival time at the workstation and the estimated picking time.

[0277] In some embodiments, at least one container to be transported may contain containers that have been taken away by the second transport device and containers that have not yet been taken away by the second transport device. The first device can determine the time when the second transport device carries the containers to be transported to the workstation and the time when it leaves the workstation, depending on the situation and based on different data.

[0278] Step 1403: Construct a queuing timeline based on the arrival time and departure time of each of the at least one second handling devices at the workstation.

[0279] In some embodiments, the first device can determine the dwell time period for the second handling device to arrive at the workstation and perform picking based on the arrival time and departure time. By determining the dwell time period, the first device can arrange at least one second handling device at the workstation in the order of arrival to obtain a queuing timeline.

[0280] For example, the RMS system constructs a timeline (including arrival and departure times) of the workstation queuing for the shelves based on the distance between the workstation and the shelf, the robot, the order content, the robot's average operating speed, and the estimated picking time.

[0281] In the above embodiments, by periodically updating the estimated picking time, the estimated picking time used by the first device in the process of building a queuing timeline for the workstation can be more accurate, thereby making the constructed queuing timeline more precise.

[0282] In some embodiments, a queuing timeline can be constructed based on the arrival time and departure time of the workstation by determining the order in which at least one second handling device arrives at the workstation.

[0283] In some embodiments, the sorting value of at least one second handling device arriving at the workstation can be determined by arranging at least one second handling device according to the order in which the second handling devices arrive at the workstation.

[0284] In some embodiments, the dwell time period is the time period between the arrival time and departure time of the second handling equipment at the workstation.

[0285] In some embodiments, the first device can arrange at least one second handling device arriving at the workstation in chronological order based on arrival and departure times to construct a queuing timeline for at least one second handling device at the workstation.

[0286] For example, as shown in Figures 19A and 19B, which are schematic diagrams of the timeline construction, the RMS system calculates and determines the arrival time t11 and departure time t13 of shelf 1, the arrival time t21 and departure time t23 of shelf 2, the arrival time t31 and departure time t33 of shelf 3, and the arrival time t41 and departure time t43 of shelf 4, thereby constructing the queuing timeline for the workstation as shown in Figures 19A and 19B. The shelf switching time can be the time from the completion of picking on shelf 1 to the start of picking on shelf 2.

[0287] In the above embodiments, for at least one second handling device assigned to the workstation queue, the first device can construct a queuing timeline for the workstation based on whether the picking target has been obtained, the location of the second handling device and / or the location of the container to be handled, the average operating speed, and the estimated picking time, and at the same time determine the arrival time and departure time of at least one second handling device.

[0288] Figure 15 is a flowchart of a method for determining whether a workstation is in an idle state according to some embodiments of this disclosure. Based on the embodiments shown in Figures 12-14, Figure 15 further describes step 1203 in Figure 12. As shown in Figure 15, determining whether a workstation is in an idle state includes at least one of the following two cases, namely, step 1501 and / or step 1502:

[0289] Step 1501: Based on the queuing timeline, between the arrival time of the first second transport device and the departure time of the last second transport device, if there is a first idle time period that does not overlap with the dwell time period of the at least one second transport device, determine that the workstation is in an idle state during the first idle time period.

[0290] In some embodiments, the first idle time period is from the departure time of the preceding second transport device to the arrival time of the subsequent second transport device.

[0291] In some embodiments, the idle duration is the period from the start time to the end time of an idle time period. In other words, the idle duration is the length of the first idle time period.

[0292] In some embodiments, if the dwell time periods of the second transport devices with consecutive sorting values ​​in at least one second transport device do not overlap, it can be determined that the workstation is in an idle state if there is a first idle time period between the arrival time of the first second transport device and the departure time of the last second transport device that does not overlap with the dwell time period of each of the at least one second transport device, that is, if there is no second transport device queuing at the workstation during the first idle time period, it means that there is an idle time period without second transport devices queuing in the queuing timeline of the workstation, and thus it is determined that the workstation is in an idle state during the first idle time period.

[0293] Step 1502: Based on the queuing timeline, within the dwell time of the first target transport device in at least one second transport device, if there is a second idle time period that does not overlap with the dwell time period of other second transport devices, determine that the workstation is in an idle state during the second idle time period.

[0294] In some embodiments, the idle time of the workstation is the period from the start time to the end time of the idle time period. In other words, the idle time is the duration of the second idle time period.

[0295] In some embodiments, if there is a second idle time period within the dwell time period of the first target transport device in at least one second transport device that does not overlap with the dwell time periods of other second transport devices, it can be determined that the workstation is in an idle state. This means that if there is a second idle time period within the dwell time period of the first target transport device that does not overlap with the dwell time periods of other second transport devices besides the first target transport device, it means that there is a situation in the current queuing timeline where only one second transport device is queuing at the workstation during the second idle time period, thereby determining that the workstation is in an idle state during the second idle time period.

[0296] In some embodiments, determining whether a workstation is in an idle state can be that the workstation has a first idle time period during which no second handling equipment is queued, and / or there are multiple idle time periods during which only one second handling equipment is queued, indicating that the workstation is in an idle state.

[0297] For example, the RMS system considers a workstation to be in a queue state if at least two or more shelves are in the arrival state at the same time, and an idle state if there is only one shelf or no shelf in the queue.

[0298] For example, Figure 19A illustrates the construction of a queuing timeline to determine idle periods. If a workstation has only one shelf in queue between time t33 and t43, it is determined to be idle during this time period. Furthermore, if there are no shelves in queue during the workstation idle time period after t43, the workstation is also determined to be idle.

[0299] In the above embodiment, the first device constructs a queuing timeline and determines the idle time period when the workstation is in an idle state based on the queuing timeline and the idle state judgment logic. It then replenishes the workstation with handling equipment according to the idle time. Compared with the current logic for judging the idle state, the idle time period is extended, so that the RMS system can also replenish the second handling equipment during the extended idle time period, thus avoiding the workstation not being replenished with the second handling equipment in a timely manner.

[0300] Figure 16 is a flowchart of a method for supplementing a second handling device according to some embodiments of this disclosure. Based on the embodiments shown in Figures 12-15, Figure 16 further describes step 1204 in Figure 12, as shown in Figure 16, which includes the following steps 1601-1603:

[0301] Step 1601: Obtain first status information of at least one candidate container and second status information of at least one candidate handling device.

[0302] In some embodiments, the first status information includes whether a second handling device has been assigned, the current location of the second handling device if it has been assigned, or the current location of the second handling device if it has not been assigned. The second status information includes whether a handling task has been executed, the end location of the handling task, and / or the current location of the handling task if it has not been executed.

[0303] In some embodiments, at least one candidate container is a container to be transported corresponding to at least one second order among a plurality of pick orders.

[0304] In some embodiments, the status of the candidate transport device can be either performing a transport task or not performing a transport task. The second status information obtained includes the end position of the transport task being performed or the current position of the transport task not yet being performed.

[0305] In some embodiments, the status of the candidate container includes whether a second handling device has been assigned or not, and the first status information includes the current location of the assigned second handling device or the current location of the unassigned second handling device.

[0306] Step 1602: Based on the first state information, the second state information, the average operating speed of the candidate handling equipment, and the estimated picking time, determine the first arrival time and the first departure time of at least one candidate handling equipment.

[0307] In some embodiments, for a candidate picking target assigned to a second handling device, the first arrival time of the candidate handling device after carrying the corresponding candidate container to the workstation and the first departure time of the candidate handling device from the workstation can be determined based on the current position of the candidate container, the current position of the assigned candidate handling device, the average operating speed of the candidate handling device, and the expected picking time.

[0308] In some embodiments, for candidate containers that have not been assigned a second handling device, a nearby candidate handling device that is currently performing a handling task can be assigned to it. After the candidate handling device that is currently performing the handling task has completed its task, the candidate container is then transported to the workstation. Based on the current position of the candidate container, the end position of the candidate handling device that is currently performing the handling task, the average operating speed of the candidate handling device, and the estimated picking time, the first arrival time and the first departure time of the candidate handling device after completing its task and carrying the corresponding candidate container can be determined.

[0309] In some embodiments, for a candidate picking target that has not been assigned a second handling device, a candidate handling device that has not yet performed a handling task can be assigned to it, and the candidate handling device will transport the candidate container to the workstation. Then, based on the current position of the candidate handling device that has not yet performed a handling task, the position of the candidate container, the average operating speed of the candidate handling device, and the estimated picking time, the first arrival time and the first departure time of the candidate handling device can be determined.

[0310] Step 1603: Based on the first arrival time and the first departure time, determine at least one additional handling device among at least one candidate handling device.

[0311] In some embodiments, determining at least one additional handling device from at least one candidate handling device can be achieved through the following methods:

[0312] Option 1:

[0313] In some embodiments, step 1603, determining at least one supplementary transport device among at least one candidate transport device, includes: determining a candidate ranking value for the arrival of at least one candidate transport device at the workstation based on a first arrival time; determining the candidate transport device corresponding to a first time unit whose first arrival time is before the start time of the idle period and whose first departure time is before the end time of the idle period as the first supplementary transport device; and determining the first supplementary transport device as the supplementary transport device if the dwell time of the first supplementary transport device is greater than the idle period.

[0314] In some embodiments, determining the candidate ranking value for the arrival of at least one candidate handling device at the workstation based on the first arrival time may be determining the order in which the at least one candidate handling device arrives at the workstation.

[0315] In some embodiments, the idle time period may be the first idle time period and / or the second idle time period determined in FIG15.

[0316] For example, as shown in the schematic diagram of the replenishment shelves in Figure 19B, the workstation is in an idle state during the time period from t23 when shelf 2 leaves to t31 when shelf 3 arrives. The idle time period is t23-t31. The arrival times of candidates H1, H2, H3, and H4 to the workstation are determined from the candidate shelves, as shown in Figure 19B. The horizontal coordinate corresponding to the left side of the rectangle of each candidate shelf is the arrival time of the candidate shelf, and the horizontal coordinate corresponding to the right side of the rectangle is the departure time of the candidate shelf. Candidates H1 and H2 have the same arrival time, and candidates H2 and H3 have the same departure time. The candidate shelves are sorted according to their arrival time to obtain the priority order of the candidate shelves. When the arrival times of H1 and H2 are the same, the priority order can be based on the longer dwell time, that is, H3 > H2 > H1 > H4. However, the departure times of H1, H2, H3, and H4 are all after the end time of the idle time period. Therefore, it is necessary to re-determine the replenishment shelves from the candidate shelves.

[0317] In some embodiments, the candidate transport device corresponding to the first time unit whose first arrival time is before the start time of the idle period and whose first departure time is before the end time of the idle period is determined as the first transport device to be supplemented. This can be done by determining the candidate transport device that can arrive before the start of the idle state of the workstation and leave before the arrival of the next second transport device as the first transport device to be supplemented.

[0318] In some embodiments, when the dwell time of the first transport device to be replenished is longer than the idle time period, the first transport device to be replenished can be identified as a transport device to be replenished if it can leave the workstation within a first time unit before the end of the idle time period. In other words, if the dwell time of the first transport device to be replenished can fill the idle time period of the workstation and will not prevent the next queued transport device to arrive at the workstation for picking, the first transport device to be replenished is identified as a transport device to be replenished.

[0319] In some embodiments, if the dwell time of the first transport device to be replenished is less than the idle time period, a first candidate ranking value of the first transport device to be replenished is determined among at least one candidate transport device; at least one candidate transport device whose candidate ranking value is greater than or equal to the first candidate ranking value and whose first departure time is before the end time of the idle time period is determined as at least one second transport device to be replenished; and at least one second transport device to be replenished whose dwell time periods are consecutive among the at least one second transport device to be replenished is determined as at least one transport device to be replenished.

[0320] In some embodiments, if the dwell time of the first equipment to be replenished is less than the idle time, it may be that the dwell time of the first equipment to be replenished cannot fill the idle time of the workstation, resulting in an idle time that still needs to be replenished after the first equipment to be replenished leaves the workstation.

[0321] In some embodiments, determining the first candidate ranking value of the first to be supplemented transport device among at least one candidate transport device may be a ranking value for determining the arrival of the first to be supplemented transport device at the workstation.

[0322] In some embodiments, at least one candidate transport device whose candidate ranking value is greater than or equal to the first candidate ranking value and whose departure time is before the end of the idle period is determined as at least one second transport device to be supplemented. This can be a candidate transport device that leaves before the end of the idle period and whose arrival time is after the first transport device to be supplemented, and the first transport device to be supplemented is determined as at least one second transport device to be supplemented.

[0323] In some embodiments, determining a second candidate shelf with consecutive dwell time periods as at least one replenishable handling device can be done by determining a second replenishable handling device whose departure time coincides with or is close to (e.g., less than t seconds, where t can be a preset value) the replenishment time period among the at least one second replenishable handling device. In other words, multiple candidate handling devices with consecutive dwell time periods within the idle time are determined as at least one replenishable handling device.

[0324] [Corrected according to Rule 91, 26.11.2025] For example, the fastest-arriving shelf is prioritized for replenishment from the candidate shelves, i.e., a time-priority scheduling strategy is used to schedule candidate shelves to replenish idle time periods. Candidate shelves that leave after the end of the idle time period are excluded, resulting in replenished shelves. For the schematic diagram of scheduling candidate shelves shown in Figure 19B, H3 is excluded because its departure time is greater than the end time of the idle time period, and the same applies to H2. In the above scheme one, candidate shelves 1, 2, 3, and 4 do not meet the scheduling strategy; therefore, shelves H1, H2, H3, and H4 will not be determined as replenished shelves.

[0325] Option 2:

[0326] In some embodiments, determining at least one supplementary transport device among at least one candidate transport device based on a first arrival time and a first departure time includes: determining a candidate transport device whose dwell time is within an idle time period as at least one third supplementary transport device based on the dwell time period of the at least one candidate transport device; and determining a third supplementary transport device that meets a first condition as at least one supplementary transport device.

[0327] In some embodiments, the idle time period may be the first idle time period and / or the second idle time period determined in FIG5.

[0328] In some embodiments, determining a candidate transport device whose dwell time is within an idle period as at least one third transport device to be supplemented may be done by determining at least one third transport device to be supplemented from at least one candidate transport device according to the principle that the second transport device arriving at the workstation is the most numerous.

[0329] In some embodiments, the first condition may be that the dwell time periods do not overlap and / or the dwell time period is less than T. Then, determining a third replenishing transport device that meets the first condition as at least one replenishing transport device may involve determining a third replenishing transport device whose dwell time periods do not overlap, or a third replenishing transport device whose dwell time period is less than T, or a third replenishing transport device whose dwell time periods do not overlap and are less than T. Wherein, T may be a pre-set minimum dwell time period that can meet the expected picking time.

[0330] For example, if the picking time of candidate shelves cannot cover the length of the idle time period, the RMS system can schedule robots to move candidate shelves to the workstation according to the scheduling strategy of delivering the most shelves to the workstation per unit time.

[0331] In the above embodiments, based on the start and end times of the idle time period when the first workstation is in an idle state, at least one material handling device to be supplemented can be determined from at least one candidate material handling device according to different supplementation strategies, so as to ensure that the second material handling device in the queue of the workstation is not interrupted and improve the operating efficiency of the workstation.

[0332] Figure 17 is a flowchart of another device scheduling method provided by some embodiments of this disclosure. Based on the embodiments shown in Figures 12-16, as shown in Figure 17, it includes the following steps 1701-1702:

[0333] Step 1701: Determine the average dwell time of each workstation based on the dwell time of each workstation in at least one workstation on at least one second handling device within a first time period.

[0334] In some embodiments, the first time period can be a pre-set time period, or any time period before the current time, such as one hour.

[0335] In some embodiments, based on the dwell time of at least one second handling device at each workstation within a first time period prior to the current time, historical data within the first time period may be obtained, and the average dwell time of each workstation may be determined based on the actual dwell time of each second handling device at the workstation within the first time period.

[0336] For example, add a scheduled task to calculate the average shelf dwell time of the workstation over a recent period (the most recent hour).

[0337] Step 1702: For the first workstation whose average dwell time meets the second condition, determine the first number of additional handling equipment.

[0338] In some embodiments, the first quantity may be a preset value, or a different value may be set according to different second conditions, or a different value may be set according to different workstations, and this disclosure does not limit this.

[0339] In some embodiments, the second condition may be that the average dwell time is less than the first duration, where the first duration may be a custom duration. This indicates that the first workstation that meets the second condition has high processing efficiency or fast picking speed.

[0340] In some embodiments, the second condition may be that at least one workstation, sorted by average dwell time from longest to shortest, has a sorting value greater than M or the last N digits of the sorting value. This indicates that the first workstation satisfying the second condition has high processing efficiency or fast picking speed.

[0341] In the above embodiments, determining a first workstation that meets the second condition can identify at least one workstation with high processing efficiency or fast picking speed, thereby adding a fixed number of second handling devices to that workstation for queuing, so as to give full play to the high efficiency of that workstation.

[0342] Figure 18 is a flowchart of determining additional handling equipment according to some embodiments of this disclosure. Based on the embodiments shown in Figures 12 and 17, as shown in Figure 18, it includes the following steps 1801-1803:

[0343] Step 1801: After assigning at least one second handling device to the first workstation for queuing, obtain first status information of at least one candidate picking target and second status information of at least one candidate handling device.

[0344] In some embodiments, the first status information includes whether a second handling device has been assigned, the current location of the second handling device if it has been assigned, or the current location of the second handling device if it has not been assigned. The second status information includes whether a handling task has been executed, the end location of the handling task, and / or the current location of the handling task if it has not been executed.

[0345] In some embodiments, at least one candidate picking target is the picking target corresponding to at least one third order among a plurality of orders to be picked.

[0346] In some embodiments, after assigning at least one second handling device to the first workstation in the queue, if the average dwell time of the first workstation in the first time period before the current time meets the second condition, then the first status information of at least one candidate picking target and the second status information of at least one candidate handling device are obtained.

[0347] In some embodiments, the status of the candidate transport device can be either performing a transport task or not performing a transport task. The second status information obtained includes the end position of the transport task being performed or the current position of the transport task not yet being performed.

[0348] In some embodiments, the status of the candidate picking target includes whether it has been assigned to a second handling device or not, and the first status information includes the current location of the assigned second handling device or the current location of the unassigned second handling device.

[0349] Step 1802: Based on the first state information, the second state information, and the average operating speed of the candidate transport equipment, determine the arrival ranking value of at least one candidate transport equipment to the first workstation.

[0350] In some embodiments, for a candidate picking target assigned to a second handling device, the arrival time of the candidate handling device after carrying the corresponding candidate container to the first workstation can be determined based on the current position of the candidate container, the current position of the assigned candidate handling device, and the average operating speed.

[0351] In some embodiments, for a candidate container that has not been assigned a second handling device, the arrival time of the candidate handling device carrying the corresponding candidate picking target to the first workstation can be determined based on the current position of the candidate container, the current position of the newly assigned candidate handling device, and the average operating speed.

[0352] In the above embodiments, at least one candidate transport device is sorted based on its arrival time at the first workstation to obtain an arrival sorting value for each candidate transport device.

[0353] Step 1803: Candidate handling equipment whose arrival ranking value is less than or equal to the first number is determined as supplementary handling equipment for the first workstation.

[0354] In some embodiments, the first quantity may be a pre-set number of supplementary handling devices. The first quantity may be set to different values ​​based on the average dwell time of the workstation; in other words, multiple first quantities may be set, and this disclosure does not limit this.

[0355] In some embodiments, determining a candidate handling device that arrives at a sorting value less than or equal to a first number as a supplementary handling device for the first workstation may be determining a fixed number (the first number) of candidate handling devices that arrive at the first workstation first as supplementary handling devices.

[0356] For example, all workstations are ranked from longest to shortest. The workstations ranked Nth are given a separate queue number H. The increased queue number is not counted in the replenishment shelf quantity, nor is the preset queue number threshold of the workstations in the system changed.

[0357] In the above embodiments, by adding a timed task, namely monitoring the average dwell time of the workstation within a preset time period, the actual picking efficiency of the workstation is measured, and a certain number of queuing second handling devices are allocated to workstations with high picking efficiency, thereby further enhancing the efficiency of the workstation.

[0358] The following is a specific implementation of a queuing method provided in this disclosure, where the picking target is a shelf. It includes the following steps:

[0359] 1. Build a queuing timeline for the workstation's queuing shelves and determine the idle time periods.

[0360] The logic for constructing the queuing timeline is as follows: RMS calculates the time it takes for the robot to reach the shelf based on the robot's current position, the shelf's position, and the robot's average speed; then, based on the distance between the shelf and the workstation and the robot's average speed, it calculates the time it takes for the shelf to reach the workstation, i.e., the arrival time; and based on the estimated picking time and the arrival time, it calculates the time it takes to leave the workstation, i.e., the departure time.

[0361] If the robot has already picked up the shelf, RMS calculates the arrival time based on the distance between the shelf and the workstation and the robot's average speed.

[0362] The logic for determining idle time periods is as follows: a workstation is considered not to be idle only if at least two or more shelves are in the arrival state at the same time. If there is only one shelf or no shelf in the queue at this time, that is, the number of shelves in the queue is less than 2, then the workstation is in an idle state.

[0363] Figure 19A illustrates the construction of the queuing timeline. Based on this timeline, it can be determined that the workstation is idle from time t33 to t43. Compared to the current method for determining idle periods, this extends the idle period, allowing the RMS system to perform shelf replenishment logic during the extended idle period, thus preventing untimely shelf replenishment at the workstation. The switching time is the time it takes to switch to the next shelf for picking after the previous shelf has left.

[0364] 2. Replenish shelves when the workstation is idle.

[0365] Increase the number of queued shelves at workstations during their off-peak hours and select which shelves to replenish.

[0366] When workstations experience idle periods, the fastest-reaching shelves need to be prioritized for replenishment from the candidate shelves. The RMS system acquires candidate shelf data and performs replenishment logic based on this data to determine the order (priority) of shelf replenishment. Candidate shelves and already assigned queued shelves are determined by the RMS system based on all currently processed orders, and each shelf is considered equivalent. Therefore, shelves scheduled earlier can pick orders first, and shelves scheduled later can pick orders later; the order of order picking is irrelevant.

[0367] Figure 19B shows a schematic diagram of the scheduling candidate shelves, where the scheduling priority order is H3>H2>H1>H4.

[0368] Scheduling logic: Based on the shelf location, average robot speed, and average picking time, the arrival and departure times of the shelves are estimated. Based on the idle time of the workstation, the arrival and departure times of the shelves, and the scheduling principles, the priority order of shelf scheduling is determined.

[0369] The scheduling strategy can be arrival time priority or departure time priority.

[0370] Scheduling principle: Try to ensure that candidate shelves arrive before the start time of the idle period and leave before the end time of the idle period; or, if the picking time of candidate shelves cannot cover the entire idle period, multiple candidate shelves can be selected so that the picking time of the combination of multiple candidate shelves covers the idle period.

[0371] Since the RMS system can obtain data for each shelf (location, quantity of goods in the included orders), it can determine the priority order of candidate shelves based on the data, ensuring that the maximum number of shelves delivered to the workstation per unit time is reached, and that the shelves delivered to the workstation by the robot are continuously flowing, thereby improving the workstation picking efficiency.

[0372] 3. Calculate the average dwell time at workstations and dynamically increase the number of queues.

[0373] Dwell time definition: the time when the shelf arrives at the workstation — the time when the shelf leaves the workstation.

[0374] 3.1 Add a scheduled task to calculate the average shelf dwell time of the workstation over a recent period (the most recent hour).

[0375] 3.2. Rank all workstations according to their average dwell time (from longest to shortest). The workstations ranked Nth (indicating high processing efficiency and fast picking speed) will have an additional queue length M.

[0376] 3.3 The increased queue size is not included in the dynamically increased queue size; that is, it is only considered as a temporary increase in queue size and does not change the preset queue size of workstations in the system.

[0377] Figure 20 shows a schematic diagram of a device scheduling apparatus provided in some embodiments of the present disclosure. As shown in Figure 20, the device scheduling apparatus 2000 may include:

[0378] The acquisition module 2001 is configured to acquire the operating data of multiple first handling devices and the task data of multiple tasks to be executed, wherein the multiple first handling devices include at least one device to be scheduled.

[0379] The processing module 2002 is configured to process the running data and task data to obtain feature data of multiple first handling devices and multiple tasks to be executed.

[0380] Analysis module 2003 is configured to use a pre-trained time prediction model to analyze feature data and determine the task duration for the device to be scheduled to execute the task.

[0381] The determination module 2004 is configured to determine the target task among multiple tasks to be executed based on the task duration of the task to be executed by the device to be scheduled, so as to schedule the device to be scheduled to execute the target task.

[0382] In some embodiments, the multiple tasks to be executed include multiple retrieval tasks, each retrieval task being used to retrieve a container to be transported, and the task duration including a first task duration for the scheduled device to execute the retrieval task; the determining module 2004 is configured to determine the retrieval task with the shortest first task duration among the multiple retrieval tasks as the target task based on the first task duration for each retrieval task executed by the scheduled device; and to schedule the scheduled device to execute the target task to retrieve the container to be transported at the container location corresponding to the target task.

[0383] In some embodiments, the multiple tasks to be executed include multiple container delivery tasks, which are used to move containers to be transported on the scheduling device to any workstation. The task duration includes a second task duration for the scheduling device to execute the corresponding container delivery task. The determining module 2004 is configured to, when multiple container delivery tasks indicate that the containers to be transported are to be moved to the same workstation, obtain a first remaining task execution duration for the same workstation; determine a target task among the multiple container delivery tasks based on the second task duration for each scheduling device to execute the corresponding container delivery task and the first remaining task execution duration for the same workstation; and schedule the scheduling device to execute the target task so that the scheduling device moves the containers to be transported to the corresponding workstation.

[0384] In some embodiments, the determining module 2004 is configured to determine a first time difference for each delivery task based on the second task duration of each scheduled device executing the corresponding delivery task and the first remaining task execution duration of the same workstation; and to determine the delivery task with the smallest first time difference as the target task.

[0385] In some embodiments, the multiple tasks to be executed include multiple container retrieval tasks and container delivery tasks corresponding to each container retrieval task. The container retrieval task is used to retrieve the container to be transported, and the container delivery task is used to transport the container to be transported to the workstation. The task duration includes a third task duration for the scheduled device to execute the container retrieval task and a fourth task duration for executing the container delivery task. The determining module 2004 is configured to determine the total task duration for the scheduled device to execute each task based on the third task duration for the scheduled device to execute each container retrieval task and the fourth task duration for the container delivery task corresponding to each container retrieval task; determine the target task among the multiple tasks to be executed based on the total task duration for the scheduled device to execute each task and the second remaining task execution duration of the workstation corresponding to each task to be executed; and schedule the scheduled device to execute the target task to retrieve the container to be transported at the container location corresponding to the target task and transport the container to be transported to the corresponding workstation.

[0386] In some embodiments, the determining module 2004 is configured to, when each delivery task indicates that the container to be transported is to be transported to the same workstation, calculate a second time difference between the total task duration of the scheduled device executing each task to be executed and the second remaining task execution duration of the workstation corresponding to each task to be executed; and determine the task to be executed with the smallest second time difference as the target task.

[0387] In some embodiments, the determining module 2004 is configured to, when each delivery task indicates that the container to be transported is to be transported to any workstation, calculate a third time difference between the total task duration of the scheduled device executing each task to be executed and the second remaining task execution duration of the workstation corresponding to each task to be executed; if there is only one task to be executed with a third time difference less than a preset threshold, determine the task to be executed with a third time difference less than the preset threshold as the target task; if there are multiple tasks to be executed with a third time difference less than the preset threshold, determine the task to be executed with the shortest total task duration among the multiple tasks to be executed with a third time difference less than the preset threshold as the target task.

[0388] In some embodiments, the pre-trained time prediction model includes a first time prediction model and a second time prediction model, and the feature data includes first feature data and second feature data. The analysis module 2003 is configured to use the first time prediction model to analyze the first feature data and determine the third task duration for the device to be scheduled to perform the box retrieval task; and to use the second time prediction model to analyze the second feature data and determine the fourth task duration for the device to be scheduled to perform the box delivery task.

[0389] In some embodiments, the operation data includes equipment operation data of the first transport device and task execution data of the first transport device currently performing a task; the equipment operation data includes at least one of the scheduling status, travel path, location information, and travel speed of the first transport device, and the task execution data includes the task information and / or task execution status of the first transport device currently performing a task; the task data includes at least one of the starting position, destination position, task information, and task type of the task to be executed.

[0390] In some embodiments, the processing module 2002 is configured to generate, based on running data and task data, device feature data of a plurality of first handling devices, path feature data of the devices to be scheduled to reach the destination locations of each task to be executed, and task feature data of the other handling devices among the plurality of first handling devices except for at least one device to be scheduled, currently executing tasks; and generate feature data based on device feature data, path feature data and task feature data.

[0391] In some embodiments, the equipment feature data includes at least one of the scheduling status, travel path, location information, and travel speed of the first handling equipment; the path feature data includes the travel path of the equipment to be scheduled to reach the execution location of each task to be executed, and at least one of the equipment heat statistics and the number of feature points in the travel path; the task feature data includes the task statistics of other tasks currently being executed by the first handling equipment.

[0392] In some embodiments, the acquisition module 2001 is further configured to acquire information about at least one second handling device queuing at the workstation, the second handling device being a device that carries the picking target and transports it to the workstation. The determination module 2004 is further configured to determine, based on the information about the at least one second handling device, the arrival time of the at least one second handling device at the workstation and the departure time of the at least one second handling device from the workstation.

[0393] The equipment scheduling device 2000 may further include a judgment module and a supplement module. The judgment module is configured to determine whether the workstation is in an idle state based on the arrival and departure times of at least one second handling device. The supplement module is configured to supplement the workstation with a second handling device according to the idle duration if the workstation is in an idle state.

[0394] In some embodiments, the determining module 2004 is further configured to acquire location information, the average operating speed of the second handling equipment, and the estimated picking time. The location information includes the location of at least one second handling equipment and / or the location of at least one picking target. Based on the location information, the average operating speed, and the estimated picking time, a queuing timeline is constructed for at least one second handling equipment at the workstation, and the arrival time and departure time of at least one second handling equipment at the workstation are determined. The queuing timeline includes the sorting value of at least one second handling equipment arriving at the workstation, the arrival time, the departure time, and the dwell time period.

[0395] In some embodiments, the determining module 2004 is further configured to determine the arrival time of each second handling device at the workstation based on the location information and average operating speed of each of the at least one second handling devices; determine the departure time of each second handling device from the workstation based on the arrival time and the estimated picking time; and construct a queuing timeline based on the arrival time and departure time of each of the at least one second handling devices.

[0396] In some embodiments, the determining module 2004 is further configured to determine the time when the second transport device arrives at the workstation based on the position of the container to be transported and the average operating speed when the position of the second transport device is the same as the position of the container to be transported; and to determine the time when the second transport device arrives at the workstation based on the position of the second transport device, the position of the container to be transported, and the average operating speed when the position of the second transport device is different from the position of the container to be transported.

[0397] In some embodiments, the determining module is configured to, based on the queuing timeline, determine that the workstation is idle during a first idle time period, between the arrival time of the first second transport device and the departure time of the last second transport device among at least one second transport device, provided that there is a first idle time period that does not overlap with the dwell time period of at least one second transport device. The first idle time period is from the departure time of the preceding second transport device to the arrival time of the subsequent second transport device. And / or, based on the queuing timeline, determine that the workstation is idle during a second idle time period, between the dwell time period of the first target transport device among at least one second transport device, provided that there is a second idle time period that does not overlap with the dwell time period of other second transport devices. The idle duration is from the start time to the end time of the idle time period.

[0398] In some embodiments, the supplementary module is further configured to acquire first status information of at least one candidate container and second status information of at least one candidate handling device. The first status information includes whether a second handling device has been assigned, the current location of the assigned second handling device, or the current location of the unassigned second handling device. The second status information includes whether a handling task has been executed, the end location of the executed handling task, and / or the current location of the unexecuted handling task. Based on the first status information, the second status information, the average operating speed of the candidate handling devices, and the estimated picking time, the module determines a first arrival time and a first departure time of the at least one candidate handling device. Based on the first arrival time and the first departure time, the module determines at least one handling device to be supplemented among the at least one candidate handling device.

[0399] In some embodiments, the supplementation module is further configured to determine, based on the first arrival time, a candidate sorting value for at least one candidate transport device arriving at the workstation; to determine the candidate transport device whose first arrival time is before the start time of the idle period and whose first departure time is before the end time of the idle period as the first transport device to be supplemented; and to determine the first transport device to be supplemented as the transport device to be supplemented if the dwell time of the first transport device to be supplemented is longer than the idle period.

[0400] In some embodiments, the supplementation module is further configured to, when the dwell time of the first supplemented transport device is less than the idle time period, determine a first candidate ranking value of the first supplemented transport device among at least one candidate transport device; determine at least one candidate transport device whose candidate ranking value is greater than or equal to the first candidate ranking value and whose first departure time is before the end time of the idle time period as at least one second supplemented transport device; and determine at least one second supplemented transport device among the at least one second supplemented transport devices whose dwell time periods are consecutive as at least one supplemented transport device.

[0401] In some embodiments, the supplementation module is further configured to determine candidate handling devices whose dwell time is within an idle time period as at least one third handling device to be supplemented, based on the dwell time period of at least one candidate handling device; and to determine a third handling device to be supplemented that meets a first condition as at least one handling device to be supplemented, the first condition including that the dwell time periods do not overlap and / or the dwell time period is less than the minimum value of the dwell time period that meets the expected picking time.

[0402] In some embodiments, the supplementary module is further configured to determine the average dwell time of each workstation based on the dwell time of each workstation on at least one second handling device within a first time period; and to determine a first number of supplementary handling devices for a first workstation whose average dwell time satisfies a second condition, the second condition including that the average dwell time is less than the first duration.

[0403] In some embodiments, the supplementary module is further configured to, after assigning at least one second handling device to the first workstation in the queue, acquire first status information of at least one candidate container and second status information of at least one candidate handling device, the first status information including whether a second handling device has been assigned, the current position of an assigned second handling device or the current position of an unassigned second handling device, and the second status information including whether a handling task has been executed, the end position of an executed handling task and / or the current position of an unexecuted handling task; determine an arrival ranking value for at least one candidate handling device arriving at the first workstation based on the first status information, the second status information, and the average operating speed of the candidate handling devices; and determine candidate handling devices whose arrival ranking value is less than or equal to a first number as supplementary handling devices for the first workstation.

[0404] In the above embodiments, the first device can determine the arrival time and departure time of each queued second handling device at the workstation based on the information of the queued second handling devices assigned to each workstation, so as to determine the idle time of the workstation in an idle state. Based on the idle time, the first device can replenish the workstation with second handling devices in a timely manner to avoid the interruption of the queued second handling devices and achieve efficient operation of the workstation.

[0405] Figure 21 shows a schematic diagram of the structure of an electronic device provided in an embodiment of the present disclosure. The specific embodiments of the present disclosure do not limit the specific implementation of the electronic device.

[0406] As shown in Figure 21, the electronic device can be the control device or the first device in the above embodiments. Specifically, the electronic device may include: a processor 2102, a communications interface 2104, a memory 2106, and a communications bus 2108.

[0407] The processor 2102, communication interface 2104, and memory 2106 communicate with each other via communication bus 2108. Communication interface 2104 is used to communicate with other network elements such as clients or other servers. The processor 2102 executes program 2110, specifically performing the relevant steps described above in the embodiments of the device scheduling method and / or model training method.

[0408] Specifically, program 2110 may include program code, which includes computer-executable instructions.

[0409] The processor 2102 may be a central processing unit (CPU), an application-specific integrated circuit (ASIC), or one or more integrated circuits configured to implement embodiments of the present disclosure. The electronic device may include one or more processors of the same type, such as one or more CPUs; or it may include processors of different types, such as one or more CPUs and one or more ASICs.

[0410] Memory 2106 is used to store program 2110. Memory 2106 may include high-speed RAM memory, and may also include non-volatile memory, such as at least one disk storage device.

[0411] Specifically, program 2110 can be called by processor 2102 to cause the electronic device to execute the steps in the above-mentioned device scheduling method and / or model training method.

[0412] This disclosure also provides a computer-readable storage medium storing at least one executable instruction that, when executed on an electronic device / device scheduling system, causes the electronic device / device scheduling system to perform the device scheduling method and / or model training method in any of the above method embodiments.

[0413] The executable instructions can be used to cause the electronic device / device scheduling system to perform the steps in the above-mentioned device scheduling method and / or model training method.

[0414] Embodiments of this disclosure also provide a computer program product, including a computer program that is executed by a processor using the device scheduling method and / or model training method described in the above embodiments of this disclosure.

[0415] The algorithms or displays provided herein are not inherently related to any particular computer, virtual system, or other device. Furthermore, the embodiments disclosed herein are not directed to any particular programming language.

[0416] Numerous specific details are set forth in the specification provided herein. However, it will be understood that embodiments of this disclosure may be practiced without these specific details. Similarly, for the sake of brevity and to aid in understanding one or more aspects of the invention, in the foregoing description of exemplary embodiments of this disclosure, various features of embodiments of this disclosure are sometimes grouped together in a single embodiment, figure, or description thereof. The claims, which follow the detailed description, are hereby expressly incorporated into that detailed description, wherein each claim itself is a separate embodiment of this disclosure.

[0417] Those skilled in the art will understand that the modules in the device of the embodiment can be adaptively changed and placed in one or more devices different from that embodiment. Modules, units, or components in the embodiment can be combined into a single module, unit, or component, and further, they can be divided into multiple sub-modules, sub-units, or sub-components, except that at least some of such features and / or processes or units are mutually exclusive.

[0418] It should be noted that the above embodiments are illustrative of this disclosure and not restrictive, and that those skilled in the art can devise alternative embodiments without departing from the scope of the appended claims. In the claims, any reference signs placed between parentheses should not be construed as limiting the claims. The word "comprising" does not exclude the presence of elements or steps not listed in the claims. The word "a" or "an" preceding an element does not exclude the presence of a plurality of such elements. This disclosure can be implemented by means of hardware comprising several different elements and by means of a suitably programmed computer. In the unit claims enumerating several means, several of these means may be embodied by the same item of hardware. The use of the words first, second, and third, etc., does not indicate any order. These words can be interpreted as names. The steps in the above embodiments, unless otherwise specified, should not be construed as limiting the order of execution.

Claims

1. A device scheduling method, the method comprising: Acquire operational data of multiple first handling devices and task data of multiple tasks to be executed, wherein the multiple first handling devices include at least one device to be scheduled; The operation data and the task data are processed to obtain the feature data of the plurality of first handling devices and the plurality of tasks to be executed; A pre-trained time prediction model is used to analyze the feature data to determine the task duration for the device to be scheduled to execute the task. Based on the task duration of the scheduled device in executing the task to be executed, a target task is determined among the multiple tasks to be executed, so as to schedule the scheduled device to execute the target task.

2. The method according to claim 1, wherein, The plurality of tasks to be executed includes a plurality of box retrieval tasks, wherein the box retrieval task is used to retrieve the container to be transported, and the task duration includes the first task duration for the scheduled device to execute the box retrieval task; The step of determining a target task from among the multiple tasks to be executed based on the task duration of the task to be executed by the device to be scheduled, and then scheduling the device to be scheduled to execute the target task, includes: Based on the first task duration of each of the box retrieval tasks executed by the scheduled device, the box retrieval task with the shortest first task duration among the multiple box retrieval tasks is determined as the target task. The device to be scheduled is dispatched to perform the target task in order to retrieve the container to be transported from the container location corresponding to the target task.

3. The method according to claim 1, wherein, The multiple tasks to be executed include multiple box delivery tasks. The box delivery task is used to move the container to be transported on the scheduling device to any workstation. The task duration includes the second task duration for the scheduling device to execute the corresponding box delivery task. The step of determining a target task from among the multiple tasks to be executed based on the task duration of the task to be executed by the device to be scheduled, and then scheduling the device to be scheduled to execute the target task, includes: When multiple delivery tasks instruct the container to be transported to the same workstation, obtain the first remaining task execution time of the same workstation. Based on the second task duration of each scheduled device executing the corresponding box delivery task and the first remaining task execution duration of the same workstation, the target task is determined among the multiple box delivery tasks. The scheduled device is instructed to perform the target task so that it can move the container to be transported to the corresponding workstation.

4. The method according to claim 3, wherein, The step of determining the target task from the plurality of box delivery tasks based on the second task duration of each of the scheduled devices executing the corresponding box delivery task and the first remaining task execution duration of the same workstation includes: Based on the second task duration of each scheduled device executing the corresponding box delivery task and the first remaining task execution duration of the same workstation, a first time difference is determined for each box delivery task. The task with the smallest time difference is identified as the target task.

5. The method according to claim 1, wherein, The multiple tasks to be executed include multiple box retrieval tasks and box delivery tasks corresponding to each box retrieval task. The box retrieval task is used to retrieve the container to be transported, and the box delivery task is used to transport the container to be transported to the workstation. The task duration includes the third task duration for the scheduled device to execute the box retrieval task and the fourth task duration for executing the box delivery task. The step of determining a target task from among the multiple tasks to be executed based on the task duration of the task to be executed by the device to be scheduled, and then scheduling the device to be scheduled to execute the target task, includes: Based on the third task duration of the device to be scheduled executing each of the box retrieval tasks and the fourth task duration of the box delivery task corresponding to each of the box retrieval tasks, the total task duration of the device to be scheduled executing each of the tasks to be executed is determined. Based on the total task duration of the scheduled device executing each of the tasks to be executed and the second remaining task execution duration of the workstation corresponding to each task to be executed, the target task is determined from the plurality of tasks to be executed; The device to be scheduled is dispatched to perform the target task, so as to retrieve the container to be transported from the container location corresponding to the target task and transport the container to be transported to the corresponding workstation.

6. The method according to claim 5, wherein, The step of determining the target task from the plurality of tasks to be executed based on the total task duration of the scheduled device executing each of the tasks to be executed and the second remaining task execution duration of the workstation corresponding to each task to be executed includes: When each of the delivery task instructions moves the container to be transported to the same workstation, calculate the second time difference between the total task duration of the scheduled equipment executing each of the tasks to be executed and the second remaining task execution duration of the workstation corresponding to each task to be executed. The task to be executed with the smallest second time difference is determined as the target task.

7. The method according to claim 5, wherein, The step of determining the target task from the plurality of tasks to be executed based on the total task duration of the scheduled device executing each of the tasks to be executed and the second remaining task execution duration of the workstation corresponding to each task to be executed includes: When each of the aforementioned delivery task instructions moves the container to be transported to any workstation, calculate a third time difference between the total task duration of the scheduled equipment executing each of the aforementioned tasks and the second remaining task execution duration of the workstation corresponding to each of the aforementioned tasks. If there is only one task with a third time difference less than a preset threshold, then the task with the third time difference less than the preset threshold is identified as the target task. If there are multiple tasks with a third time difference less than a preset threshold, the task with the shortest total task duration among these multiple tasks with a third time difference less than the preset threshold is determined as the target task.

8. The method according to claim 5, wherein, The pre-trained time prediction model includes a first time prediction model and a second time prediction model. The feature data includes first feature data and second feature data. The step of using the pre-trained time prediction model to analyze the feature data and determine the task duration for the device to be scheduled to execute the task includes: Using the first time prediction model, the first feature data is analyzed to determine the third task duration for the device to be scheduled to perform the box retrieval task; The second time prediction model is used to analyze the second feature data to determine the fourth task duration for the scheduled device to perform the box delivery task.

9. The method according to any one of claims 1-8, wherein, The operational data includes the equipment operation data of the first handling equipment and the task execution data of the first handling equipment currently performing a task; the equipment operation data includes at least one of the scheduling status, travel path, location information, and travel speed of the first handling equipment, and the task execution data includes the task information and / or task execution status of the first handling equipment currently performing a task. The task data includes at least one of the following: the start position, destination position, task information, and task type of the task to be executed.

10. The method according to any one of claims 1-8, wherein, The process of processing the operational data and the task data to obtain feature data of the plurality of first handling devices and the plurality of tasks to be executed includes: Based on the operational data and the task data, the following are generated: equipment feature data of the plurality of first handling devices, path feature data of the devices to be scheduled to reach the destination locations of each of the tasks to be executed, and task feature data of the other handling devices among the plurality of first handling devices, excluding the at least one device to be scheduled, currently executing tasks. The feature data is generated based on the device feature data, the path feature data, and the task feature data.

11. The method according to claim 10, wherein, The equipment feature data includes at least one of the following: the scheduling status, travel path, location information, and travel speed of the first handling equipment; The path feature data includes the travel path of the scheduled device to the execution location of each of the tasks to be executed, and at least one of the following: device heat statistics and the number of feature points in the travel path; The task feature data includes task statistics information of the other first handling equipment currently performing tasks.

12. The method according to claim 1, wherein, The scheduling of the device to be scheduled to execute the target task includes: At least one second handling device is dispatched to move the containers to be handled to the workstation for queuing, wherein the devices to be dispatched include the at least one second handling device; After scheduling the device to be scheduled to execute the target task, the method further includes: Obtain information about at least one second handling device that is queued at the workstation, wherein the second handling device is a device that carries a container to be handled and moves it to the workstation, and the device to be scheduled includes the second handling device; Based on the information of the at least one second handling device, determine the arrival time of the at least one second handling device at the workstation and the departure time from the workstation; Based on the arrival time and departure time of the at least one second handling device, determine whether the workstation is in an idle state; When the workstation is in an idle state, replenish the workstation with handling equipment according to the idle time.

13. The method according to claim 12, wherein, Determining the arrival time of the at least one second handling device at the workstation and its departure time from the workstation based on the information of the at least one second handling device includes: Acquire location information, average operating speed of the second handling equipment, and estimated picking time, wherein the location information includes the location of the at least one second handling equipment and / or the location of at least one container to be handled; Based on the location information, the average operating speed, and the estimated picking time, a queuing timeline is constructed for at least one second handling device at the workstation to determine the arrival time of the at least one second handling device at the workstation and the departure time of the at least one second handling device from the workstation. The queuing timeline includes the order value of the at least one second handling device arriving at the workstation, the arrival time, the departure time, and the dwell time period.

14. The method according to claim 13, wherein, The step of constructing a queuing timeline for at least one second handling device at the workstation based on the location information, the average operating speed, and the estimated picking time includes: Based on the location information corresponding to each of the at least one second handling devices and the average operating speed, the arrival time of each second handling device at the workstation is determined. Based on the arrival time at the workstation and the estimated picking time, the departure time of each second handling device from the workstation is determined; The queuing timeline is constructed based on the arrival time and departure time of each of the at least one second handling devices at the workstation.

15. The method according to claim 14, wherein, The determination of the arrival time of each second transport device at the workstation based on the location information corresponding to each of the at least one second transport device and the average operating speed includes at least one of the following: When the position of the second handling device is the same as the position of the container to be handled, the time for the second handling device to arrive at the workstation is determined based on the position of the container to be handled and the average operating speed. If the location of the second handling device is different from the location of the container to be handled, the arrival time of the second handling device at the workstation is determined based on the location of the second handling device, the location of the container to be handled, and the average operating speed.

16. The method according to claim 15, wherein, Determining whether the workstation is in an idle state based on the arrival time and departure time of the at least one second handling device includes: Based on the queuing timeline, if there is a first idle time period between the arrival time of the first second transport device and the departure time of the last second transport device among the at least one second transport device, and this first idle time period does not overlap with the dwell time period of the at least one second transport device, it is determined that the workstation is in the idle state during the first idle time period, where the first idle time period is from the departure time of the preceding second transport device to the arrival time of the following second transport device; and / or Based on the queuing timeline, during the dwell time of the first target transport device in the at least one second transport device, if there is a second idle time period that does not overlap with the dwell time period of other second transport devices, it is determined that the workstation is in the idle state during the second idle time period, and the idle duration is from the start time to the end time of the idle time period.

17. The method according to any one of claims 12-16, wherein, The step of replenishing the workstation with handling equipment based on the idle time when the workstation is in the idle state includes: Acquire first status information of at least one candidate container and second status information of at least one candidate transport device. The first status information includes whether a second transport device has been assigned, the current position of the assigned second transport device, or the current position of the unassigned second transport device. The second status information includes whether a transport task has been executed, the end position of the executed transport task, and / or the current position of the unexecuted transport task. Based on the first status information, the second status information, the average operating speed of the candidate handling equipment, and the estimated picking time, the first arrival time and the first departure time of the at least one candidate handling equipment are determined. Based on the first arrival time and the first departure time, at least one additional transport device is determined from the at least one candidate transport device.

18. The method according to claim 17, wherein, The step of determining at least one additional transport device from the at least one candidate transport device based on the first arrival time and the first departure time includes: Based on the first arrival time, determine the candidate ranking value for the arrival of the at least one candidate handling device at the workstation; The candidate handling equipment corresponding to the first arrival time before the start time of the idle period and the first departure time before the end time of the idle period is determined as the first handling equipment to be supplemented. If the dwell time of the first equipment to be replenished is longer than the idle time period, the first equipment to be replenished is identified as the equipment to be replenished.

19. The method according to claim 18, further comprising: If the dwell time of the first replenished transport equipment is less than the idle time, a first candidate ranking value of the first replenished transport equipment among the at least one candidate transport equipment is determined; At least one candidate handling device whose candidate sorting value is greater than or equal to the first candidate sorting value and whose first departure time is before the end time of the idle period is identified as at least one second handling device to be supplemented. The second transport equipment that has a continuous dwell time among the at least one second transport equipment to be supplemented is identified as the at least one transport equipment to be supplemented.

20. The method of claim 17, wherein, The step of determining at least one additional transport device from the at least one candidate transport device based on the first arrival time and the first departure time includes: Based on the dwell time of the at least one candidate handling equipment, the candidate handling equipment whose dwell time falls within the idle time period is identified as at least one third handling equipment to be supplemented; The third piece of handling equipment that meets the first condition is identified as the at least one piece of handling equipment to be supplemented. The first condition includes non-overlapping dwell time periods and / or dwell time periods less than the minimum dwell time period required to meet the expected picking time.

21. The method according to claim 12, further comprising: The average dwell time of each workstation is determined based on the dwell time of each workstation in at least one workstation within a first time period at at least one second handling device. For a first workstation whose average dwell time meets the second condition, a first number of supplementary handling equipment are determined, wherein the second condition includes an average dwell time less than a first duration.

22. The method according to claim 21, wherein, For the first workstation whose average dwell time meets the second condition, determining a first number of supplementary handling devices includes: After assigning at least one second handling device to the first workstation in the queue, first status information of at least one candidate container and second status information of at least one candidate handling device are obtained. The first status information includes whether a handling device has been assigned, the current position of the assigned handling device, or the current position of the unassigned handling device. The second status information includes whether a handling task has been executed, the end position of the executed handling task, and / or the current position of the unexecuted handling task. Based on the first status information, the second status information, and the average operating speed of the candidate handling equipment, the arrival ranking value of the at least one candidate handling equipment to the first workstation is determined. Candidate handling devices whose arrival sort value is less than or equal to the first number are identified as supplementary handling devices for the first workstation.

23. A model training method, the method comprising: Obtain operational sample data of multiple first handling devices and task sample data of multiple handling tasks, wherein the task sample data includes the actual task duration of the first handling device executing the corresponding handling task; The running sample data and the task sample data are processed to obtain training data for the plurality of first handling devices and the plurality of handling tasks; The training data is input into the initial time prediction model to obtain the predicted task duration for the first handling device to perform the handling task; A pre-trained time prediction model is obtained by training based on the predicted task duration and the actual task duration.

24. The method according to claim 23, wherein, The task sample data also includes the start and end positions of the first handling equipment performing the corresponding handling task. Before processing the running sample data and the task sample data to obtain the training data for the plurality of first handling equipment and the plurality of handling tasks, the method further includes: Based on at least one of the starting position, the ending position, and the actual task duration of the first handling device performing the corresponding handling task, determine whether the sample data of the first handling device performing the corresponding handling task is abnormal data; If the sample data is abnormal, delete the sample data. The sample data that is considered abnormal includes any of the following: The distance between the starting position and the ending position is greater than a first distance threshold, and the actual task time for the first handling device to perform the corresponding handling task is less than a first time threshold. The distance between the starting position and the ending position is less than the second distance threshold, and the actual task duration of the first handling device performing the corresponding handling task is greater than the second time threshold. The starting position is the same as the ending position; The distance between the starting position and the ending position is less than a third distance threshold; The actual task duration for the first handling device to perform the corresponding handling task is less than the third time threshold. The actual task duration for the first handling device to perform the corresponding handling task is greater than the fourth time threshold.

25. The method according to claim 23, wherein, The initial time prediction model includes any one or more of the following: random forest algorithm model, LightGBM algorithm model, and linear regression algorithm model.

26. An equipment scheduling system, the system comprising: A plurality of first handling devices, wherein the plurality of first handling devices includes at least one device to be scheduled; The control device is configured as follows: Obtain the operating data of the plurality of first handling devices and the task data of the plurality of tasks to be executed; The operation data and the task data are processed to obtain the feature data of the plurality of first handling devices and the plurality of tasks to be executed; A pre-trained time prediction model is used to analyze the feature data to determine the task duration for the device to be scheduled to execute the task. Based on the task duration of the scheduled device in executing the task to be executed, a target task is determined among the multiple tasks to be executed, so as to schedule the scheduled device to execute the target task.

27. An electronic device comprising: processor; as well as Memory for storing the executable instructions of the processor; The processor is configured to perform the device scheduling method and / or model training operations as described in any one of claims 1-25 by executing the executable instructions.

28. A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the device scheduling method and / or model training method according to any one of claims 1-25.

Images