A multi-robot scheduling method and device for a warehouse system and related medium

Abstract

The application discloses a kind of warehousing system multi-robot scheduling method, device and related medium, the method includes obtaining storage site-robot associated data set and statistics warehouse material quantity and sequencing, obtain warehouse site allocation data;Using warehouse site allocation data analysis warehouse demand, obtain the warehouse site allocation data;Wave is arranged based on the warehouse site allocation data, and tray loading instruction set is obtained;According to tray loading instruction set, extract the task set of line tray to be thrown and sequencing, generate line scheduling instruction set according to priority;Tray loading instruction set and line instruction set are accessed to platform processing, and scheduling instruction data is output according to tray state.The application is calculated by tray loading instruction set and the line instruction set calculated by being accessed to platform processing, and scheduling instruction data is output according to tray state, so that multi-job robot can be unified scheduling optimization, and overall task completion time is reduced.

Description

Technical Field

[0001] This invention relates to the field of logistics and distribution technology, and in particular to a multi-robot scheduling method, device and related media for a warehousing system. Background Technology

[0002] In warehousing operations, there are often multiple batches and different types of workers simultaneously handling inbound and outbound shipments and conveyor belt operations. Furthermore, there are numerous robots available on-site, the work areas are dispersed, and there are differences in queuing and occupancy status at platforms and workstations. Existing technologies typically rely on human experience or static rules to assign and distribute tasks. This makes it difficult to uniformly optimize the order of tasks, assignment targets, and workstation access during peak periods. This can easily lead to uneven task distribution, a coexistence of idle and congested robots, and disordered queuing at key workstations, resulting in extended overall task completion times and making it difficult to complete all tasks within a short period. Summary of the Invention

[0003] This invention provides a method, apparatus, and related medium for scheduling multiple robots in a warehousing system, aiming to solve the technical problem in the prior art where it is difficult to perform unified scheduling and optimization of multiple robots, resulting in excessively long overall task completion time.

[0004] In a first aspect, embodiments of the present invention provide a multi-robot scheduling method for a warehousing system, comprising:

[0005] The robot coordinates, storage location coordinates, and area identifiers are obtained respectively to calculate the correspondence between the lanes and obtain the storage location-robot association dataset;

[0006] Based on the location-robot association dataset, the quantity of incoming materials is counted and sorted to obtain the inbound location allocation data;

[0007] The outbound demand is parsed using the inbound storage location allocation data and evenly mapped to the robot area to obtain the outbound storage location allocation data;

[0008] Based on the outbound storage location allocation data, wave sorting is performed to issue pallet loading instructions in the pallet loading area according to priority, thereby obtaining a pallet loading instruction set;

[0009] Extract and sort the set of pallet tasks to be deployed according to the pallet loading instruction set, and generate a deployment scheduling instruction set according to priority.

[0010] The pallet loading instruction set and the line deployment instruction set are processed for station access, and scheduling instruction data is output according to the pallet status.

[0011] Secondly, embodiments of the present invention provide a multi-robot scheduling device for a warehousing system, comprising:

[0012] The data acquisition unit is used to acquire robot coordinates, cargo location coordinates, and area identifiers respectively, in order to calculate the correspondence of the lanes and obtain the cargo location-robot association dataset;

[0013] The material allocation unit is used to count and sort the quantity of materials entering the warehouse based on the location-robot association dataset to obtain the warehouse location allocation data;

[0014] The data parsing unit is used to parse outbound demand using the inbound storage location allocation data and evenly map it to the robot area to obtain outbound storage location allocation data.

[0015] The instruction orchestration unit is used to orchestrate waves based on the outbound storage location allocation data, so as to issue palletizing instructions in the palletizing area according to priority, thereby obtaining a palletizing instruction set;

[0016] The data sorting unit is used to extract and sort the set of pallet tasks to be deployed according to the pallet loading instruction set, and generate a deployment scheduling instruction set according to priority.

[0017] The scheduling output unit is used to perform station access processing on the pallet loading instruction set and the line deployment instruction set, and output scheduling instruction data according to the pallet status.

[0018] Thirdly, embodiments of the present invention provide a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the multi-robot scheduling method of the warehousing system of the first aspect.

[0019] Fourthly, embodiments of the present invention provide a computer-readable storage medium, wherein a computer program is stored on the computer-readable storage medium, and when the computer program is executed by a processor, it implements the multi-robot scheduling method of the warehousing system of the first aspect.

[0020] This invention provides a multi-robot scheduling method for a warehousing system, including acquiring robot coordinates, storage location coordinates, and area identifiers to calculate the correspondence between aisles and obtain a storage location-robot association dataset; counting and sorting the quantity of incoming materials based on the storage location-robot association dataset to obtain inbound storage location allocation data; parsing outbound demand using the inbound storage location allocation data and mapping it evenly to robot areas to obtain outbound storage location allocation data; arranging waves based on the outbound storage location allocation data to issue palletizing instructions in the palletizing area according to priority, obtaining a palletizing instruction set; extracting and sorting a set of pallet tasks to be deployed according to the palletizing instruction set, and generating a deployment scheduling instruction set according to priority; performing station access processing on the palletizing instruction set and the deployment instruction set, and outputting scheduling instruction data according to the pallet status. This invention enables unified scheduling optimization of multiple robots by performing station access processing on the calculated palletizing instruction set and the calculated deployment instruction set, and outputting scheduling instruction data according to the pallet status, thereby reducing the overall task completion time.

[0021] This invention also provides a multi-robot scheduling device, computer equipment, and storage medium for a warehousing system, which have the same beneficial effects as described above. Attached Figure Description

[0022] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0023] Figure 1 A flowchart illustrating a multi-robot scheduling method for a warehousing system provided in an embodiment of the present invention;

[0024] Figure 2 This is a schematic block diagram of a multi-robot scheduling device for a warehousing system provided in an embodiment of the present invention. Detailed Implementation

[0025] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0026] It should be understood that, when used in this specification and the appended claims, the terms "comprising" and "including" indicate the presence of the described features, integrals, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or collections thereof.

[0027] It should also be understood that the terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to limit the invention. As used in this specification and the appended claims, the singular forms “a,” “an,” and “the” are intended to include the plural forms unless the context clearly indicates otherwise.

[0028] It should also be further understood that the term "and / or" as used in this specification and the appended claims refers to any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0029] Please see below. Figure 1 , Figure 1 The flowchart of a multi-robot scheduling method for a warehousing system provided in an embodiment of the present invention specifically includes steps S101 to S106.

[0030] S101. Obtain robot coordinates, cargo location coordinates, and area identifiers respectively to calculate the correspondence of the aisles and obtain the cargo location-robot association dataset;

[0031] S102. Calculate and sort the quantity of materials entering the warehouse based on the warehouse location-robot association dataset to obtain warehouse location allocation data;

[0032] S103. Analyze the outbound demand using the inbound storage location allocation data and evenly map it to the robot area to obtain outbound storage location allocation data;

[0033] S104. Based on the outbound storage location allocation data, arrange the wave to issue palletizing instructions in the palletizing area according to priority, and obtain a palletizing instruction set;

[0034] S105. Extract and sort the set of pallet tasks to be deployed according to the pallet loading instruction set, and generate a deployment scheduling instruction set according to priority.

[0035] S106. Perform station access processing on the pallet loading instruction set and the line deployment instruction set, and output scheduling instruction data according to the pallet status.

[0036] In step S101, the robot coordinates of each working robot, the location coordinates of each storage location, and the area identifier are obtained respectively. The coordinate axis is selected as the aisle determination dimension according to the area identifier. By calculating the coaxial matching relationship between the robot coordinates and the storage location coordinates, the aisle correspondence between the storage location and the robot is determined, and then the storage location-robot association dataset is constructed.

[0037] In one embodiment, step S101 includes:

[0038] Based on the area identifier, the difference between the longitudinal coordinate of the robot coordinate and the longitudinal coordinate of the cargo location is calculated, and the longitudinal coordinate axis is determined to be equal when the absolute value of the difference is not greater than a first preset value, so as to determine that the current cargo location and the current robot have the same aisle, and thus obtain the longitudinal aisle data.

[0039] Based on the area identifier, the difference between the horizontal coordinate of the robot coordinate and the horizontal coordinate of the storage location is calculated, and the horizontal coordinate axis is determined to be equal when the absolute value of the difference is not greater than a second preset value, so as to determine that the current storage location and the current robot have the same aisle, and the horizontal aisle data is obtained.

[0040] The longitudinal aisle data and the transverse aisle data are aggregated to obtain the location-robot association dataset.

[0041] In this embodiment, in warehousing operation scenarios with large workloads and numerous robots, a correspondence is first established between robots and floor storage locations. Inbound materials are then evenly distributed according to robot dimensions to maintain a relatively balanced inventory of each material in front of each robot, providing a data foundation for outbound storage location allocation. Specifically, the scheduling system can read the robot coordinates of the current robot, the storage location coordinates of the current location, and the area identifier, and determine the coordinate axis used for aisle determination based on the area identifier. When the tray area is used as the current area identifier, the scheduling system uses the vertical coordinate of the robot coordinates as the y-axis coordinate value 'a' and the vertical coordinate of the storage location coordinates as the y-axis coordinate value 'b', performs difference calculation, and takes the absolute value. When |ab| <= 3, the vertical coordinate axes are considered equal, thus determining that the current storage location and the current robot share the same aisle. This aisle determination result, along with the corresponding robot identifier, storage location identifier, and area identifier, is written into the vertical aisle data to describe the aisle correspondence between storage locations and robots within the tray area. When the full-case picking area is used as the current area identifier, the scheduling system uses the horizontal coordinate of the robot coordinate as the x-axis coordinate value 'a' and the horizontal coordinate of the storage location coordinate as the x-axis coordinate value 'b'. It performs difference calculation and takes the absolute value. When |ab|<=3, it is determined that the horizontal coordinate axes are equal, and then it is determined that the current storage location and the current robot's aisle are the same. The aisle determination result, along with the corresponding robot identifier, storage location identifier, and area identifier, is written into the horizontal aisle data to describe the aisle correspondence between storage locations and robots in the full-case picking area.

[0042] After completing the calculation and recording of the vertical and horizontal aisle data, the scheduling system summarizes the two types of aisle data, integrates the reachable storage locations corresponding to the same robot under different area identifiers, and outputs a storage location-robot association dataset. In this association dataset, the scheduling system can record the material quantity statistics and sorting items of the current inventory task along the aisle dimension, so that the materials are sorted in ascending order of quantity when entering the warehouse, giving priority to aisles with fewer materials, and selecting the nearest empty storage location in the selected aisle. This allows subsequent task allocation and scheduling to have usable aisle correspondence and priority information at the data level.

[0043] In step S102, the quantity of materials corresponding to the inbound task is counted according to the aisle dimension based on the location-robot association dataset, and the quantity of materials in each aisle is sorted so that aisles with fewer materials have priority in participating in the selection of inbound storage locations. An empty storage location is selected nearby in the target aisle to obtain the inbound storage location allocation data.

[0044] In step S103, the outbound demand is analyzed using the inbound storage location allocation data, and a balanced allocation is performed within the storage location areas divided by the robot, so that the outbound tasks are more evenly distributed in the robot aisle dimension, and each robot can prioritize obtaining executable storage location tasks from its own aisle, thus obtaining the outbound storage location allocation data.

[0045] In one embodiment, step S103 includes:

[0046] The outbound demand is read using the inbound location allocation data to obtain the outbound demand of the target material, and the outbound demand is represented as the number of outbound pallets corresponding to the target material to obtain the material outbound demand data.

[0047] The material outbound demand data and the location-robot association dataset are used to calculate the outbound location allocation, and the allocation is performed evenly in the robot aisle dimension according to the divided location areas to obtain aisle location allocation data.

[0048] Based on the aisle location allocation data, task dispatch rules are generated, and tasks are dispatched from the respective aisles of the robots according to the task dispatch rules to obtain outbound location allocation data.

[0049] In this embodiment, the outbound demand is read using the inbound storage location allocation data. The outbound demand for the target material is obtained from the outbound task list, and this outbound demand is uniformly converted into a quantity expression at the pallet level to obtain the material outbound demand data. For example, when the target material is material a and the demand is x pallets, material a and x pallets are written into the material outbound demand data as input for subsequent allocation calculations. Then, the material outbound demand data is used to perform outbound storage location allocation calculations with the storage location-robot association dataset: on the one hand, according to the robot's division of storage location areas, all available outbound storage locations are assigned to the corresponding storage location areas according to the robot's operation coverage; on the other hand, equal allocation is performed at the robot aisle level, splitting the x pallet outbound task corresponding to material a into several parts and allocating them to the corresponding storage location areas in multiple robot aisles, so that the storage location allocation remains balanced among the aisles, avoiding the concentration of outbound storage locations for the same material in a few robot aisles. Based on this evenly distributed allocation result, the scheduling system integrates the aisle location allocation data, which includes at least: target material identifier, aisle identifier, corresponding robot identifier, number of outbound pallets allocated to that aisle, and the specific set of locations that accommodate that number of pallets. Through this method, the scheduling system provides the prerequisite for shorter total travel distances for the handling trolleys during the allocation phase.

[0050] After completing the aisle location allocation data, the scheduling system generates task dispatch rules based on the aisle location allocation data, and dispatches tasks from the respective aisles of the robots based on the task dispatch rules: Specifically, for each robot, the scheduling system selects outbound locations from the location set of its corresponding aisle and generates task entries, so that the outbound tasks received by each robot can be initiated within its own aisle, minimizing the extra running distance of cross-aisle retrieval and handling trolleys; when there is a non-divisible remaining quantity of pallets in the aisles, the scheduling system adds the remaining pallets to the aisles with available locations according to the preset allocation order, updates the corresponding aisle location allocation data, and outputs outbound location allocation data for subsequent scheduling execution.

[0051] In one embodiment, step S103 further includes:

[0052] Data extraction is performed on the outbound storage location allocation data to obtain pallet loading task data;

[0053] The scheduling priority is calculated using the pallet loading task data, and then sorted in descending order by the number of task types and ascending order by the difference in vertical coordinates to obtain the pallet loading scheduling instruction;

[0054] After palletizing the palletizing scheduling instruction, the whole box delivery task is determined. If the determination condition is met, the robot is controlled to go directly to the delivery station; otherwise, the goods are controlled to return to the warehouse to the nearest storage location, and the delivery scheduling task is integrated.

[0055] In this embodiment, after completing the outbound storage location allocation, the system also performs task extraction, priority calculation, and destination determination for tray loading waves. The scheduling system first extracts data from the outbound storage location allocation data, parsing each pallet task assigned to each storage location area and requiring tray loading area robots to handle. This involves extracting pallet identifiers, target material identifiers, storage location identifiers, corresponding aisle identifiers, and task type sets. Pallet tasks including full-case delivery are then included in the tray loading task pool, resulting in tray loading task data. This data carries task entries for tray loading tasks of various types, including full-case delivery, and records the number of pallet task types and pallet coordinates for each task entry to support subsequent sorting and dispatching. After obtaining the tray loading task data, the scheduling system calculates the scheduling priority using this data. Specifically, the scheduling system, targeting robots in the tray loading area, first reads the y-axis coordinate value from the robot coordinate system of each robot in the tray loading area, and then reads the y-axis coordinate value corresponding to each pallet or its target location, calculating the difference between the y-axis and the robot coordinates. Next, using the number of pallet task types as the primary sorting field and the difference between the y-axis and the robot coordinates as the secondary sorting field, a candidate execution sequence is generated, ordered in descending order by the number of pallet task types and in ascending order by the difference between the y-axis and the robot coordinates, thus obtaining the priority order of the pallet loading scheduling instructions. To avoid task accumulation at a single robot workstation, before issuing the pallet loading scheduling instruction, the scheduling system also reads the current number of tasks at the robot workstation and in the queue. Only when the total number of tasks at the robot workstation and in the queue is less than the robot's upper limit is the pallet task allowed to be written into the corresponding robot's pallet loading scheduling instruction. The robot upper limit is configured after on-site physical location survey and is used to limit the number of tasks that the same robot can handle in parallel.

[0056] After the pallet loading scheduling command is issued and the pallet loading is completed, the scheduling system executes the pallet loading scheduling command and then determines the full-case delivery task to determine the subsequent destination of the pallet and integrate it into the delivery scheduling task. The scheduling system uses the pallet loading robot as the current execution subject. When it detects that the pallet loading status of the corresponding pallet task is completed, it determines whether the pallet still contains a full-case delivery task. At the same time, it performs idle detection and load detection on the full-case delivery robot undertaking the full-case delivery operation. Among them, the idle detection corresponds to the determination node in the flowchart where there is an idle full-case delivery robot, and the load detection is used to determine whether the number of tasks assigned to the delivery robot has not reached the upper limit. If the determination result is Y, and there are still full-case delivery tasks and the number of tasks assigned to the delivery robots has not reached the upper limit, the scheduling system directly generates a direct dispatch from the tray loading robot side, dispatches the pallet task from the tray loading robot to the delivery robot's workstation, and writes the destination into the delivery dispatch task; if the determination result is N, the scheduling system generates a return-to-warehouse dispatch, returns the goods to the nearest floor storage location or the corresponding location in the floor storage area of ​​the delivery robot, and sets the pallet task status to waiting for an idle robot to be dispatched, so that the pallet remains accessible in the floor storage area until the conditions for subsequent delivery operations are met.

[0057] In step S104, tasks are arranged in waves based on outbound location allocation data, and pallet loading instructions are issued in the pallet loading area according to preset priority. Among them, pallet tasks that include full-box delivery can be selected first, and the priority is calculated by combining information such as the number of pallet task types and the robot coordinate difference. At the same time, the task issuance control is combined with the upper limit of the number of tasks at the robot workstation and the queue position, thereby obtaining the pallet loading instruction set.

[0058] In one embodiment, step S104 includes:

[0059] Based on the outbound storage location allocation data, wave tasks are extracted to distinguish between single-task pallets and multi-task pallets, and a set of wave candidate tasks is generated.

[0060] The single-task pallets are filtered using the wave candidate task set, and priority is calculated by sorting the pallet task types in ascending order and the robot lateral distance in ascending order to obtain pallet rule data;

[0061] The robot coordinates and pallet coordinates are read separately, and the coordinates are sorted according to the pallet rule data to obtain priority data;

[0062] Pallets are assigned to robots based on the priority data, and the pallet type and aisle identifier are recorded to obtain wave allocation data.

[0063] The wave allocation data is programmed with instructions, each pallet task is written into a pallet loading instruction and sent to the corresponding robot to obtain a pallet loading instruction set.

[0064] In this embodiment, the outbound location allocation data is first scanned and fields extracted. Each outbound task is aggregated by pallet dimension, and the pallet identifier, the corresponding job type set, the pallet coordinates, and the aisle identifier where the pallet is located are read. The scheduling system marks pallets containing only one job type as single-job pallets and pallets containing two or more job types as multi-job pallets. Both types of pallet tasks are written into the wave candidate task set for priority calculation and allocation within the same wave's candidate range. After generating the wave candidate task set, the scheduling system performs a filtering process on the set, extracting single-job pallets as priority scheduling objects, and calculating pallet rule data for these single-job pallets. Specifically, the scheduling system uses the number of pallet task types as the first ranking factor and the lateral difference between the robot and the pallet as the second ranking factor. The lateral difference is obtained by calculating the difference between the x-coordinate value of the robot and the x-coordinate value of the pallet and taking the absolute value. The scheduling system generates sorting rules in ascending order of the number of pallet task types and the lateral distance between robots, and writes these sorting rules into the pallet rule data. This allows the same robot to prioritize pallet tasks with single-type tasks and closer lateral distances when processing candidate pallets.

[0065] Furthermore, the scheduling system reads the coordinates of the delivery robots participating in the wave scheduling and the pallet coordinates in the wave candidate task set, and sorts the coordinates of each pallet task according to the pallet rule data, outputting priority data. For ease of explanation, let the x-coordinate value of delivery robot a be 10, the x-coordinate value of delivery robot b be 20; the x-coordinate values ​​of pallet 1 corresponding to full-box picking be 11, pallet 2 be 15, and pallet 3 be 22; the x-coordinate values ​​of pallet 4 corresponding to full-box picking or tray loading be 10, pallet 5 be 21, and pallet 6 be 20. In this example, the scheduling system calculates the lateral distance between robot a and each pallet, and combines this with an ascending order of pallet task types to determine the priority of robot a acquiring tasks from high to low: Pallet 1, Pallet 4, Pallet 2, Pallet 3, Pallet 6, Pallet 5. Similarly, the scheduling system calculates the lateral distance between robot b and each pallet, and combines this with an ascending order of pallet task types to determine the priority of robot b acquiring tasks from high to low: Pallet 3, Pallet 6, Pallet 5, Pallet 2, Pallet 1, Pallet 4. These results are used as priority data for subsequent allocation. Based on this priority data, the scheduling system performs pallet-to-robot allocation, and simultaneously records the pallet task type and aisle identifier during the allocation process to obtain wave allocation data. The scheduling system prioritizes dispatching single-task pallets according to the priority queues of each robot, and prioritizes dispatching pallets in the same aisle when available, so that tasks can be initiated in aisles reachable by the robots. In the example, the scheduling system assigns pallet 1 (single-task in the same aisle), pallet 4 (multiple-task in the same aisle), and pallet 2 (single-task in different aisles) to robot a, and assigns pallet 3 (single-task in the same aisle), pallet 6 (multiple-task in the same aisle), and pallet 5 (multiple-task in the same aisle) to robot b. The system writes the set of tasks and aisle identifier corresponding to each pallet into the wave allocation data for subsequent station access and workstation collaboration in the execution process.

[0066] After obtaining the wave allocation data, the scheduling system performs instruction orchestration on the wave allocation data, writes each pallet task into the pallet loading instruction item, and the instruction item includes at least the fields of pallet identifier, target workstation, target robot identifier, pallet coordinates, job type set and aisle identifier, and aggregates the instruction items into a pallet loading instruction set according to the robot dimension, and sends it to the corresponding robot to start wave execution.

[0067] In step S105, the set of pallet tasks to be deployed is extracted according to the pallet loading instruction set, and the tasks are sorted according to priority to generate a deployment scheduling instruction set. On the deployment robot side, single-job pallet tasks can be scheduled first, and the order of tasks is determined by combining information such as the number of pallet task types and the robot coordinate difference.

[0068] In one embodiment, step S105 includes:

[0069] The pallet loading instruction set is parsed to extract pallet task items initiated by the ground stacking area and pointing to the delivery operation, and the set of pallet tasks to be delivered is obtained.

[0070] The workstation status of the robot station is obtained according to the set of tasks to be delivered pallets, and the priority from the queuing station to the workstation is configured as the first priority, the priority from the storage area to the workstation is configured as the second priority, and the priority from the storage area to the queuing station is configured as the third priority. The station priority configuration data is then integrated.

[0071] For each line tray task corresponding to the platform priority configuration data, platform access determination is performed. If the corresponding work station status is empty, it is scheduled to the work station according to the second priority. If the corresponding work station status is not empty, it is scheduled to enter the queuing work station according to the third priority. The platform access task data is then integrated.

[0072] Based on the platform access task data, the queuing sequence is updated in the queuing station. When the work station is detected to be empty, the tasks in the queuing station are scheduled to the work station according to the first priority to obtain the queuing update data.

[0073] The queue update data is sorted and instructions are arranged. The correspondence between the storage area, queuing station, work station and robot is written into the instruction item and the line dispatching instruction set is output.

[0074] In this embodiment, the pallet loading instruction set is parsed, and the initiation area, target operation type, and pallet identification information in each instruction are read one by one. Pallet task items whose initiation area is the ground stacking area and whose target operation type points to the delivery operation are filtered out. The filtered pallet task items are further processed by extracting fields such as pallet coordinates, target station identifier, and target robot identifier, and a set of pallet tasks to be delivered is obtained. This ensures that subsequent station access determination is based solely on this set of pallet tasks to be delivered. The scheduling system obtains the workstation status of the robot station based on the set of pallet tasks to be delivered. The robot station is configured with a ground stacking area, queuing stations, and workstations. In one example structure, two sets of workstation units are used in parallel: each set of workstation units includes at least one workstation and multiple queuing stations corresponding to that workstation. After a pallet task enters the station from the ground stacking area, it first establishes an access relationship with a queuing station or a workstation, and then the workstation connects with the robot to perform the delivery operation. When reading the status of workstations, the scheduling system categorizes them into empty and non-empty states. The non-empty state is used as the trigger condition for entering a queuing workstation, ensuring that pallet tasks enter the queuing workstation to replenish the pending goods when the workstation is not empty. Based on this platform structure, the scheduling system configures platform priorities: queuing workstations to workstations are configured as the first priority, the storage area to workstations as the second priority, and the storage area to queuing workstations as the third priority, resulting in integrated platform priority configuration data. Specifically, the first priority is used to retrieve tasks from queuing workstations when a workstation is idle; the second priority is used to directly connect storage area tasks to workstations when a workstation is idle and does not require queuing; and the third priority is used to connect storage area tasks to queuing workstations when a workstation is busy.

[0075] During the platform access determination phase, the scheduling system performs a determination on each pallet task corresponding to the platform priority configuration data. For each pallet task, the scheduling system first locates its target platform and target workstation, and reads the status of the corresponding workstation. When the corresponding workstation status is empty, the scheduling system schedules the pallet task from the stacking area to the workstation according to the second priority, and writes the workstation occupancy identifier and task record. When the corresponding workstation status is not empty, the scheduling system schedules the pallet task from the stacking area to the queuing workstation according to the third priority, and in an example structure, prioritizes access to the upstream queuing workstation, and then writes its connection with the downstream queuing workstation into the task record, thereby obtaining the platform access task data. To adapt to scenarios with multiple queuing workstations, the scheduling system simultaneously maintains the queuing position number and arrival order when accessing the queuing workstation, so that subsequent queuing sequence updates have an executable data foundation. Based on the platform access task data, the scheduling system performs queuing sequence updates within the queuing workstation. Specifically, the scheduling system continuously monitors the status of workstations. When it detects that a workstation's status has changed from non-empty to empty, it schedules the first-order task in the corresponding downstream queue to the workstation according to the first priority, and updates the task's destination from the queue to the workstation. When the downstream queue is cleared, the scheduling system migrates the first-order task in the upstream queue to the downstream queue, updating the queue position number and arrival order, thereby maintaining the continuity of supply from the queue to the workstation and obtaining queue update data. For two sets of parallel workstation units, the scheduling system independently maintains their respective queue sequences and workstation statuses, ensuring that each set of workstations can obtain tasks from its corresponding queue according to the first priority, while allowing the storage area to access different workstation units according to the second or third priority.

[0076] Finally, the scheduling system sorts and orchestrates the queue update data, generating delivery scheduling instructions based on the task's current location and next-hop destination. Each delivery scheduling instruction includes at least the correspondence between the storage area, queuing station, work station, and robot, and fields such as pallet identifier, platform identifier, current station identifier, target station identifier, priority identifier, and queue number. The scheduling system sorts the instructions under the same platform according to priority and queue number, and outputs a delivery scheduling instruction set to drive pallet tasks to connect to the storage area, queuing station, and work station according to priority rules, with the robot completing the delivery operation at the work station.

[0077] In step S106, the pallet loading instruction set and the line dispatching instruction set are summarized, and the station access processing is performed: when a task arrives at the robot station, if the robot station status is not empty, the task is accessed to the queuing station; if the robot station status is empty, the task is accessed to the work station. After the task is completed, the scheduling system performs subsequent scheduling based on the pallet status. Empty pallets are scheduled to the disc tray machine and, after disc tray completion, are scheduled to the unloading platform (i.e., the station, the same below). Pallets with remaining material are scheduled back to the ground storage location. Thus, scheduling instruction data is output to drive each robot to execute.

[0078] In one embodiment, step S106 includes:

[0079] The pallet loading instruction set and the line deployment instruction set are summarized and bound together to obtain the station access task data;

[0080] The task data accessed at the platform is processed to identify the pallet task status, thereby determining the robot's task completion identifier and writing it into the task record to obtain task completion data.

[0081] The task data is processed to determine the remaining material on the pallets. If there is no remaining material, the robot is dispatched to the empty pallet stacking area. If there is remaining material, the robot is dispatched back to the ground storage area. The empty pallet stacking task data is then obtained by filtering.

[0082] Based on the empty pallet stacking task data, the receiving station scheduling process is performed according to the pallet status to obtain station scheduling task data.

[0083] The pallet is programmed with instructions based on the platform scheduling task data to output scheduling instruction data.

[0084] In this embodiment, when pallet loading and deployment operations are carried out in parallel, pallet tasks after platform access are uniformly processed, and destination scheduling is completed according to pallet status after the pallet task is completed. The scheduling system first summarizes and binds the pallet loading instruction set and the deployment instruction set, associating related instruction items under the same pallet identifier, the same target robot identifier, and the same platform identifier into the same task record, and supplementing fields such as task source area identifier, target workstation identifier, and operation type identifier to obtain platform access task data; wherein, the platform access task data is used to describe the succession relationship of pallet tasks from pallet loading related operations to deployment related operations, as well as the current access position and next hop destination of the pallet on the platform side. After the platform access task data is generated, the scheduling system performs pallet task status identification on the platform access task data to determine the robot operation completion identifier and write it into the task record to obtain operation completion data. Specifically, the scheduling system continuously receives robot execution feedback, identifies whether the task status field corresponding to the pallet identifier has switched from "in execution" to "robot operation completed", and writes the robot operation completion identifier, completion timestamp, workstation identifier and current pallet status into the task record when the switch is detected. At the same time, the pallet task is set to a state where subsequent destination determination can be performed, so as to ensure that subsequent residual material determination and station scheduling are carried out continuously on the same task record.

[0085] Furthermore, the scheduling system performs pallet material surplus determination on the completed task data. After the robot completes its task, the system reads the corresponding outbound quantity, delivery execution result, and pallet material status to determine whether the pallet has surplus material or not. If the determination is that there is surplus material, the system updates the pallet task to a return-to-warehouse task and dispatches the robot to carry the pallet back to the ground storage area. The robot then selects an available ground storage location to complete the return, records the target storage location identifier, and writes it into the task record. If the determination is that there is no surplus material, the system updates the pallet task to an empty pallet disposal task, dispatches the robot to the empty pallet stacking location, and writes the empty pallet stacking location identifier, arrival sequence number, and stacking status into the task record. This process filters out empty pallet stacking task data, ensuring that subsequent receiving station scheduling only targets tasks related to empty pallets.

[0086] After obtaining empty pallet stacking task data, the scheduling system performs receiving station scheduling processing based on the pallet status according to the empty pallet stacking task data, obtaining station scheduling task data. After the empty pallets are stacked, the scheduling system continues to identify the pallet status as usable for receiving and schedules the empty pallet task from the empty pallet stacking position to the receiving platform. In one embodiment, after the empty pallet is delivered to the designated location by the robot, the scheduling system can first schedule the empty pallet to the disc machine for disc processing, then write the disc processing completion mark into the task record, and schedule the empty pallet to the receiving platform to provide available pallet resources. Correspondingly, when the pallet is determined to have remaining material and has been scheduled back to the storage area, the station scheduling task data record is in the storage completion state and will not enter the receiving platform scheduling branch. Finally, the scheduling system performs instruction orchestration on the pallets according to the station scheduling task data to output scheduling instruction data. The instruction arrangement includes writing fields such as pallet identifier, target robot identifier, target location identifier (empty pallet stacking, disc machine, receiving platform or ground storage area), current status identifier, completion identifier and timestamp into the instruction item, and summarizing the instruction items according to the robot dimension to output scheduling instruction data, so that the robot can complete empty pallet stacking and receiving platform supply according to the pallet status, or complete the return of surplus material pallets to the warehouse.

[0087] In one business scenario, there are two types of workstations: robotic picking and robotic packing. The robotic picking station consists of eight stations, each with two picking workstations. Each workstation can handle three tasks simultaneously, with one in operation and the other two in a queue. Tasks at the picking workstations are assigned to a destination address. If the destination address is blocked, the corresponding task enters the queue. A conveyor belt is located behind each picking workstation to transport the goods picked by the robot to the destination address, which corresponds to the loading area. Floor-stacking inventory and picking tasks originate from the task's source address. The storage structure is a back-to-back double-extension layout, with a total of 3200 storage locations in a 40×80 rectangle. Each storage location is equipped with one pallet to hold the goods and match them with the picking task. Daily picking tasks typically cover approximately 2000 pallets. The sources of palletized goods include receiving and outbound from the high-bay warehouse; both sources can be mapped to the aforementioned source address to initiate tasks. The robotic packing station consists of six packing stations, each with two packing workstations. Each workstation can handle three tasks simultaneously, with one task in operation and the other two in a queue. Tasks at the packing workstations are assigned to a destination address; if the destination address is blocked, the task enters the queue. During packing operations, pallets can be switched to smaller, shallower pallets. These shallower pallets are stored in the automated warehouse and used as inventory for split-item processing, ensuring a continuous supply for subsequent split-item outbound and packing operations.

[0088] In summary, the processing capacity per unit time for tray loading operations in this application has been increased from 180 trays per hour to 300 trays per hour. The tray tasks are more rationally distributed at the aisle and workstation levels, significantly shortening the travel path of the transport trolley and reducing the risk of congestion caused by cross-traffic. At the same time, by setting up queuing workstations, the workstations continuously supply tasks to be executed, allowing the robot to maintain high-load operation on the platform side, thereby improving the overall tray loading throughput. In terms of full-case delivery operations, the delivery efficiency has been increased from 220 trays per hour and 3,000 cases per hour to 400 trays per hour and 5,500 cases per hour, respectively. This demonstrates that after the collaborative optimization of tray task sequencing, platform access, and workstation queuing management, the continuous operation capability and overall outbound cycle time of the delivery process have been significantly improved.

[0089] Combination Figure 2 As shown, Figure 2 This is a schematic block diagram of a multi-robot scheduling device for a warehousing system provided in an embodiment of the present invention. The multi-robot scheduling device 200 for the warehousing system includes:

[0090] The data acquisition unit 201 is used to acquire robot coordinates, cargo location coordinates, and area identifiers respectively, in order to calculate the correspondence of the lanes and obtain a cargo location-robot association dataset;

[0091] Material allocation unit 202 is used to count and sort the quantity of materials entering the warehouse according to the location-robot association dataset to obtain the warehouse location allocation data;

[0092] The data parsing unit 203 is used to parse outbound demand using the inbound storage location allocation data and evenly map it to the robot area to obtain outbound storage location allocation data.

[0093] The instruction orchestration unit 204 is used to orchestrate waves based on the outbound storage location allocation data, so as to issue palletizing instructions in the palletizing area according to priority, thereby obtaining a palletizing instruction set;

[0094] Data sorting unit 205 is used to extract and sort the set of pallet tasks to be deployed according to the pallet loading instruction set, and generate a deployment scheduling instruction set according to priority;

[0095] The scheduling output unit 206 is used to perform station access processing on the pallet loading instruction set and the line deployment instruction set, and output scheduling instruction data according to the pallet status.

[0096] In this embodiment, the data acquisition unit 201 acquires robot coordinates, storage location coordinates, and area identifiers to calculate the correspondence between lanes and obtain a storage location-robot association dataset; the material allocation unit 202 counts and sorts the quantity of materials entering the warehouse based on the storage location-robot association dataset to obtain storage location allocation data; the data parsing unit 203 uses the storage location allocation data to parse outbound demand and evenly maps it to robot areas to obtain outbound storage location allocation data; the instruction orchestration unit 204 orchestrates waves based on the outbound storage location allocation data to issue palletizing instructions in the palletizing area according to priority, obtaining a palletizing instruction set; the data sorting unit 205 extracts and sorts the set of pallet tasks to be deployed according to the palletizing instruction set, and generates a deployment scheduling instruction set according to priority; the scheduling output unit 206 performs station access processing on the palletizing instruction set and the deployment instruction set, and outputs scheduling instruction data according to the pallet status.

[0097] In one embodiment, the data acquisition unit 201 is specifically used for:

[0098] Based on the area identifier, the difference between the longitudinal coordinate of the robot coordinate and the longitudinal coordinate of the cargo location is calculated, and the longitudinal coordinate axis is determined to be equal when the absolute value of the difference is not greater than a first preset value, so as to determine that the current cargo location and the current robot have the same aisle, and thus obtain the longitudinal aisle data.

[0099] Based on the area identifier, the difference between the horizontal coordinate of the robot coordinate and the horizontal coordinate of the storage location is calculated, and the horizontal coordinate axis is determined to be equal when the absolute value of the difference is not greater than a second preset value, so as to determine that the current storage location and the current robot have the same aisle, and the horizontal aisle data is obtained.

[0100] The longitudinal aisle data and the transverse aisle data are aggregated to obtain the location-robot association dataset.

[0101] In one embodiment, the data parsing unit 203 is specifically used for:

[0102] The outbound demand is read using the inbound location allocation data to obtain the outbound demand of the target material, and the outbound demand is represented as the number of outbound pallets corresponding to the target material to obtain the material outbound demand data.

[0103] The material outbound demand data and the location-robot association dataset are used to calculate the outbound location allocation, and the allocation is performed evenly in the robot aisle dimension according to the divided location areas to obtain aisle location allocation data.

[0104] Based on the aisle location allocation data, task dispatch rules are generated, and tasks are dispatched from the respective aisles of the robots according to the task dispatch rules to obtain outbound location allocation data.

[0105] In one embodiment, the data parsing unit 203 is further specifically used for:

[0106] Data extraction is performed on the outbound storage location allocation data to obtain pallet loading task data;

[0107] The scheduling priority is calculated using the pallet loading task data, and then sorted in descending order by the number of task types and ascending order by the difference in vertical coordinates to obtain the pallet loading scheduling instruction;

[0108] After palletizing the palletizing scheduling instruction, the whole box delivery task is determined. If the determination condition is met, the robot is controlled to go directly to the delivery station; otherwise, the goods are controlled to return to the warehouse to the nearest storage location, and the delivery scheduling task is integrated.

[0109] In one embodiment, the instruction orchestration unit 204 is specifically used for:

[0110] Based on the outbound storage location allocation data, wave tasks are extracted to distinguish between single-task pallets and multi-task pallets, and a set of wave candidate tasks is generated.

[0111] The single-task pallets are filtered using the wave candidate task set, and priority is calculated by sorting the pallet task types in ascending order and the robot lateral distance in ascending order to obtain pallet rule data;

[0112] The robot coordinates and pallet coordinates are read separately, and the coordinates are sorted according to the pallet rule data to obtain priority data;

[0113] Pallets are assigned to robots based on the priority data, and the pallet type and aisle identifier are recorded to obtain wave allocation data.

[0114] The wave allocation data is programmed with instructions, each pallet task is written into a pallet loading instruction and sent to the corresponding robot to obtain a pallet loading instruction set.

[0115] In one embodiment, the data sorting unit 205 is specifically used for:

[0116] The pallet loading instruction set is parsed to extract pallet task items initiated by the ground stacking area and pointing to the delivery operation, and the set of pallet tasks to be delivered is obtained.

[0117] The workstation status of the robot station is obtained according to the set of tasks to be delivered pallets, and the priority from the queuing station to the workstation is configured as the first priority, the priority from the storage area to the workstation is configured as the second priority, and the priority from the storage area to the queuing station is configured as the third priority. The station priority configuration data is then integrated.

[0118] For each pallet task to be deployed corresponding to the platform priority configuration data, platform access determination is performed. If the corresponding work station status is empty, it is scheduled to the work station according to the second priority. If the corresponding work station status is not empty, it is scheduled to enter the queuing work station according to the third priority. The platform access task data is then integrated.

[0119] Based on the platform access task data, the queuing sequence is updated in the queuing station. When the work station is detected to be empty, the tasks in the queuing station are scheduled to the work station according to the first priority to obtain the queuing update data.

[0120] The queue update data is sorted and instructions are arranged. The correspondence between the storage area, queuing station, work station and robot is written into the instruction item and the line dispatching instruction set is output.

[0121] In one embodiment, the scheduling output unit 206 is specifically used for:

[0122] The pallet loading instruction set and the line deployment instruction set are summarized and bound together to obtain the station access task data;

[0123] The task data accessed at the platform is processed to identify the pallet task status, thereby determining the robot's task completion identifier and writing it into the task record to obtain task completion data.

[0124] The task data is processed to determine the remaining material on the pallets. If there is no remaining material, the robot is dispatched to the empty pallet stacking area. If there is remaining material, the robot is dispatched back to the ground storage area. The empty pallet stacking task data is then obtained by filtering.

[0125] Based on the empty pallet stacking task data, the receiving station scheduling process is performed according to the pallet status to obtain station scheduling task data.

[0126] The pallet is programmed with instructions based on the platform scheduling task data to output scheduling instruction data.

[0127] Since the embodiments of the apparatus and the embodiments of the method correspond to each other, please refer to the description of the embodiments of the method for the embodiments of the apparatus, which will not be repeated here.

[0128] This invention also provides a computer-readable storage medium storing a computer program thereon, which, when executed, can perform the steps provided in the above embodiments. The storage medium may include various media capable of storing program code, such as a USB flash drive, a portable hard drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0129] This invention also provides a computer device, which may include a memory and a processor. The memory stores a computer program, and when the processor calls the computer program in the memory, it can implement the steps provided in the above embodiments. Of course, the computer device may also include various network interfaces, a power supply, a graphics card, etc., to utilize the graphics card's performance to operate the model, such as for inference and training.

[0130] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the systems disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the descriptions are relatively simple; relevant parts can be referred to in the method section. It should be noted that those skilled in the art can make various improvements and modifications to this application without departing from the principles of this application, and these improvements and modifications also fall within the protection scope of the claims of this application.

[0131] It should also be noted that, in this specification, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

Claims

1. A multi-robot scheduling method for a warehousing system, characterized in that, include: The robot coordinates, storage location coordinates, and area identifiers are obtained respectively to calculate the correspondence between the lanes and obtain the storage location-robot association dataset; Based on the location-robot association dataset, the quantity of incoming materials is counted and sorted to obtain the inbound location allocation data; The outbound demand is parsed using the inbound storage location allocation data and evenly mapped to the robot area to obtain the outbound storage location allocation data; Based on the outbound storage location allocation data, wave sorting is performed to issue pallet loading instructions in the pallet loading area according to priority, thereby obtaining a pallet loading instruction set; Extract and sort the set of pallet tasks to be deployed according to the pallet loading instruction set, and generate a deployment scheduling instruction set according to priority. The pallet loading instruction set and the line dispatching instruction set are processed for station access, and dispatching instruction data is output according to the pallet status. The step of extracting and sorting the set of pallet tasks to be deployed according to the pallet loading instruction set, and generating a deployment scheduling instruction set according to priority, includes: parsing the instruction items of the pallet loading instruction set to extract pallet task items initiated by the stacking area and pointing to the deployment operation, and summarizing them to obtain the set of pallet tasks to be deployed; obtaining the workstation status of the robot station according to the set of pallet tasks to be deployed, and configuring the queue station to workstation as the first priority, the stacking area to workstation as the second priority, and the stacking area to queue station as the third priority, integrating them to obtain station priority configuration data; and setting up the station for each pallet task to be deployed corresponding to the station priority configuration data. Access determination: If the corresponding workstation status is empty, the task is scheduled to the workstation according to the second priority; if the corresponding workstation status is not empty, the task is scheduled to enter the queuing workstation according to the third priority, and the station access task data is integrated. Based on the station access task data, the queuing sequence is updated in the queuing workstation. When the workstation status is detected to be empty, the task in the queuing workstation is scheduled to the workstation according to the first priority, and the queuing update data is obtained. The queuing update data is sorted by task and the instructions are arranged. The correspondence between the stacking area, queuing workstation, workstation and robot is written into the instruction item, and the line deployment scheduling instruction set is output.

2. The multi-robot scheduling method for a warehousing system according to claim 1, characterized in that, The process of acquiring robot coordinates, storage location coordinates, and area identifiers to calculate the correspondence between lanes and obtain a storage location-robot association dataset includes: Based on the area identifier, the difference between the longitudinal coordinate of the robot coordinate and the longitudinal coordinate of the cargo location is calculated, and the longitudinal coordinate axis is determined to be equal when the absolute value of the difference is not greater than a first preset value, so as to determine that the current cargo location and the current robot have the same aisle, and thus obtain the longitudinal aisle data. Based on the area identifier, the difference between the horizontal coordinate of the robot coordinate and the horizontal coordinate of the storage location is calculated, and the horizontal coordinate axis is determined to be equal when the absolute value of the difference is not greater than a second preset value, so as to determine that the current storage location and the current robot have the same aisle, and the horizontal aisle data is obtained. The longitudinal aisle data and the transverse aisle data are aggregated to obtain the location-robot association dataset.

3. The multi-robot scheduling method for a warehousing system according to claim 1, characterized in that, The step of parsing outbound demand using the inbound storage location allocation data and mapping it evenly to the robot area to obtain outbound storage location allocation data includes: The outbound demand is read using the inbound location allocation data to obtain the outbound demand of the target material, and the outbound demand is represented as the number of outbound pallets corresponding to the target material to obtain the material outbound demand data. The material outbound demand data and the location-robot association dataset are used to calculate the outbound location allocation, and the allocation is performed evenly in the robot aisle dimension according to the divided location areas to obtain aisle location allocation data. Based on the aisle location allocation data, task dispatch rules are generated, and tasks are dispatched from the respective aisles of the robots according to the task dispatch rules to obtain outbound location allocation data.

4. The multi-robot scheduling method for a warehousing system according to claim 1, characterized in that, The step of parsing outbound demand using the inbound storage location allocation data and mapping it evenly to the robot area to obtain outbound storage location allocation data also includes: Data extraction is performed on the outbound storage location allocation data to obtain pallet loading task data; The scheduling priority is calculated using the pallet loading task data, and then sorted in descending order by the number of task types and ascending order by the difference in vertical coordinates to obtain the pallet loading scheduling instruction; wherein, the difference in vertical coordinates is calculated by the y-axis coordinate value of the robot and the y-axis coordinate value of the pallet. After palletizing the palletizing scheduling instruction, the whole box delivery task is determined. If the determination condition is met, the robot is controlled to go directly to the delivery station; otherwise, the goods are controlled to return to the warehouse to the nearest storage location, and the delivery scheduling task is integrated.

5. The multi-robot scheduling method for a warehousing system according to claim 1, characterized in that, The process of arranging wave orders based on the outbound storage location allocation data to issue palletizing instructions in the palletizing area according to priority results in a set of palletizing instructions, including: Based on the outbound storage location allocation data, wave tasks are extracted to distinguish between single-task pallets and multi-task pallets, and a set of wave candidate tasks is generated. The single-task pallets are filtered using the wave candidate task set, and priority is calculated by sorting the pallet task types in ascending order and the robot lateral difference in ascending order to obtain pallet rule data; wherein, the lateral difference is calculated by the x-axis coordinate value of the robot coordinate and the x-axis coordinate value of the pallet. The robot coordinates and pallet coordinates are read separately, and the coordinates are sorted according to the pallet rule data to obtain priority data; Pallets are assigned to robots based on the priority data, and the pallet type and aisle identifier are recorded to obtain wave allocation data. The wave allocation data is programmed with instructions, each pallet task is written into a pallet loading instruction and sent to the corresponding robot to obtain a pallet loading instruction set.

6. The multi-robot scheduling method for a warehousing system according to claim 1, characterized in that, The step of performing station access processing on the pallet loading instruction set and the line dispatching instruction set, and outputting dispatching instruction data according to the pallet status, includes: The pallet loading instruction set and the line dispatching instruction set are summarized and bound together to obtain the station access task data; The task data accessed at the platform is processed to identify the pallet task status, thereby determining the robot's task completion identifier and writing it into the task record to obtain task completion data. The task data is processed to determine the remaining material on the pallets. If there is no remaining material, the robot is dispatched to the empty pallet stacking area. If there is remaining material, the robot is dispatched back to the ground storage area. The empty pallet stacking task data is then filtered out. Based on the empty pallet stacking task data, the receiving station scheduling process is performed according to the pallet status to obtain station scheduling task data. The pallet is programmed with instructions based on the platform scheduling task data to output scheduling instruction data.

7. A multi-robot scheduling device for a warehousing system, characterized in that, include: The data acquisition unit is used to acquire robot coordinates, cargo location coordinates, and area identifiers respectively, in order to calculate the correspondence of the lanes and obtain the cargo location-robot association dataset; The material allocation unit is used to count and sort the quantity of materials entering the warehouse based on the location-robot association dataset to obtain the warehouse location allocation data; The data parsing unit is used to parse outbound demand using the inbound storage location allocation data and evenly map it to the robot area to obtain outbound storage location allocation data. The instruction orchestration unit is used to orchestrate waves based on the outbound storage location allocation data, so as to issue palletizing instructions in the palletizing area according to priority, thereby obtaining a palletizing instruction set; The data sorting unit is used to extract and sort the set of pallet tasks to be deployed according to the pallet loading instruction set, and generate a deployment scheduling instruction set according to priority. The scheduling output unit is used to perform station access processing on the pallet loading instruction set and the line dispatching instruction set, and output scheduling instruction data according to the pallet status. The data sorting unit is specifically used to parse the pallet loading instruction set to extract pallet task items initiated by the stacking area and pointing to the delivery operation, and to summarize them into a set of pallet tasks to be delivered; to obtain the workstation status of the robot station according to the set of pallet tasks to be delivered, and to configure the priority from the queuing station to the workstation as the first priority, the stacking area to the workstation as the second priority, and the stacking area to the queuing station as the third priority, and to integrate them to obtain station priority configuration data; to perform station access determination on each pallet task to be delivered corresponding to the station priority configuration data, if the corresponding workstation status is... If the status is empty, the task is scheduled to the work station according to the second priority. If the status of the corresponding work station is not empty, the task is scheduled to enter the queuing work station according to the third priority, and the station access task data is integrated. Based on the station access task data, the queuing sequence is updated in the queuing work station. When the work station status is detected to be empty, the task in the queuing work station is scheduled to the work station according to the first priority, and the queuing update data is obtained. The queuing update data is sorted by task and the instructions are arranged. The correspondence between the ground storage area, queuing work station, work station and robot is written into the instruction item, and the line dispatching instruction set is output.

8. A computer device, characterized in that, The system includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the multi-robot scheduling method for a warehousing system as described in any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the multi-robot scheduling method for a warehousing system as described in any one of claims 1 to 6.

Images