Goods shelf system for cross-roadway operation

Abstract

The invention provides a goods shelf system for cross-roadway operation, and relates to the technical field of warehouse logistics, the goods shelf system for cross-roadway operation comprises a climbing robot, a truss robot and a dispatching module, the truss robot is used for being arranged at the top of a plurality of goods shelf bodies arranged side by side in the first direction, and the dispatching module is used for dispatching the goods shelf bodies; in the first direction, the length direction and the vertical direction of the goods shelf body are respectively vertical pairwise; and the scheduling module is used for controlling the truss robot to grab the corresponding climbing robot according to the obtained to-be-executed task and executing the cross-roadway operation between the goods shelf bodies corresponding to the to-be-executed task. The truss robot is matched with the climbing robot to achieve cross-roadway operation, and the goods regulation and control efficiency of the goods shelf system is effectively improved.

Description

Technical Field

[0001] This invention relates to the field of warehousing and logistics technology, and more specifically, to a racking system for cross-aisle operations. Background Technology

[0002] As a core infrastructure in the field of automated warehousing and intelligent logistics, racking systems are widely used in e-commerce warehousing centers, intelligent manufacturing plant parts warehouses, pharmaceutical distribution warehouses, library automated storage and retrieval systems, and other scenarios. Through high-density storage design, they greatly improve space utilization, support the orderly storage and efficient flow of materials, and are key equipment to meet the "large capacity, fast turnover, and refined" management needs of modern logistics.

[0003] In related technologies, a fixed deployment mode of dedicated robots for single aisles is usually adopted. Each robot can only complete the picking and placing of material boxes in its assigned aisle (the rack body), resulting in low overall utilization of the rack system, unbalanced resource allocation, and affecting the efficiency of goods scheduling in the rack system. Summary of the Invention

[0004] The problem addressed by this invention is how to improve the efficiency of goods dispatching in a shelving system.

[0005] To address the aforementioned problems, this invention provides a racking system for cross-aisle operations.

[0006] In a first aspect, the present invention provides a racking system for cross-aisle operations, comprising a climbing robot, a gantry robot, and a scheduling module. The gantry robot is used to be mounted on top of multiple rack bodies arranged side by side along a first direction, wherein the first direction, the length direction of the rack body, and the vertical direction are perpendicular to each other. The scheduling module is used to control the gantry robot to grab the corresponding climbing robot according to the acquired task to be executed, and to perform cross-aisle operations between the rack bodies corresponding to the task to be executed.

[0007] Optionally, controlling the gantry robot to grasp the climbing robot according to the acquired task to be executed, and performing cross-aisle operations between the rack bodies corresponding to the task to be executed, includes: Construct a cost matrix based on all the climbing robots and the tasks to be performed; The task execution scheme with the minimum cost is obtained by using the Hungarian algorithm based on the cost matrix, wherein the task execution scheme includes a one-to-one correspondence between the climbing robot and the task to be executed; According to the task execution plan, the gantry robot is controlled to cooperate with the climbing robot to perform the cross-tunnel operation.

[0008] Optionally, the method for constructing the cost matrix includes: Based on each climbing robot and the task to be performed, the comprehensive cost of the corresponding elements in the cost matrix is ​​obtained through a preset cost relationship; The cost matrix is ​​generated by combining the costs of all the elements.

[0009] Optionally, the cost relationship satisfies: C ij =w1×D ij +w2×SOC i +w3×TQ j +w4×TC ij ; Among them, C ij For the i-th climbing robot performing the j-th task to be performed, D is the comprehensive cost of the element. ij SOC is the distance from the i-th climbing robot to the j-th task point of the task to be performed. i Let TQ be the battery level of the i-th climbing robot. j TC is the waiting time for the j-th task to be executed. ij The time for the i-th climbing robot to perform the j-th task across the tunnel is given by w1, w2, w3, and w4, which are the corresponding weighting coefficients.

[0010] Optionally, controlling the gantry robot to cooperate with the climbing robot to perform the cross-tunnel operation according to the task execution plan includes: Determine the corresponding starting shelf body and target shelf body based on the task to be executed; The climbing robot corresponding to the task to be performed is set on the corresponding climbing track of the starting shelf body; The gantry robot is controlled to cooperate with the corresponding climbing robot to perform the cross-aisle operation according to the cross-aisle movement path of the task to be performed. The cross-aisle movement path includes the original longitudinal path of the climbing robot from the picking position of the starting shelf body to the corresponding station, the cross-aisle transverse path of the gantry robot grabbing from the station of the starting shelf body to the station of the target shelf body, and the target longitudinal path of the climbing robot from the station of the target shelf body to the corresponding target position.

[0011] Optionally, the system also includes: Obtain the actual arrival time of the climbing robot at the corresponding station and the corresponding reservation time window; If the actual arrival time is within the reservation time window, then the actual arrival time is determined as the planned arrival time of the reservation time window; Otherwise, execute the preset conflict resolution strategy.

[0012] Optionally, the conflict resolution strategy includes: Obtain the available time before and after the scheduled time window; If the actual arrival time is within the range of the pre- and post-arrival free time, and the time interval between the actual arrival time and the initial planned arrival time of the reservation window is less than a preset adjustment threshold, then the actual arrival time is determined as the planned arrival time. Otherwise, the climbing robot's cross-lane operation is dynamically adjusted, wherein the dynamic adjustment includes site switching, task reallocation, and task priority preemption.

[0013] Optionally, the system also includes: Obtain the status data of the climbing robot corresponding to the task to be executed; If the status data meets the feasibility conditions for crossing the tunnel, then it is determined that the climbing robot can perform the tunnel crossing operation; Otherwise, select the climbing robot again.

[0014] Optionally, the status data includes health data and task data, wherein the health data includes corresponding battery health factors, mechanical health factors, communication health factors, and historical health factors; the step of determining that the climbing robot can perform the cross-lane operation when the status data meets the cross-lane feasibility conditions includes: The health evaluation value of the climbing robot is obtained by multiplying the battery health factor, the mechanical health factor, the communication health factor, and the historical health factor. When the health evaluation value is greater than the preset health threshold, the climbing robot will meet the schedulable requirements. If the task data corresponding to the climbing robot that meets the schedulable requirements satisfies the cross-lane feasibility condition, then it is determined that the climbing robot can perform the cross-lane operation.

[0015] Optionally, the task data includes cross-tunnel operation waiting time, cross-tunnel movement time, cross-tunnel energy consumption cost, cross-tunnel resource conflict probability, cross-tunnel benefit, task execution time in this tunnel, operation energy consumption in this tunnel, resource conflict probability in this tunnel, and task waiting time in this tunnel; the step of determining that the climbing robot can perform the cross-tunnel operation when the task data corresponding to the climbing robot that meets the schedulable requirements satisfies the cross-tunnel feasibility condition includes: Based on the cross-lane operation waiting time, the cross-lane movement time, the cross-lane energy consumption cost, the cross-lane resource conflict probability, and the cross-lane benefit, the corresponding cross-lane scheduling cost is obtained by weighted summation. Based on the task execution time of this roadway, the operation energy consumption of this roadway, the resource conflict probability of this roadway, and the task waiting time of this roadway, the corresponding scheduling cost of this roadway is obtained by weighted summation; If the product of the cross-lane scheduling cost and the preset cross-lane coefficient is greater than the current lane scheduling cost, then the climbing robot is determined to be capable of performing the cross-lane operation.

[0016] The beneficial effects of the cross-aisle racking system of the present invention are as follows: the gantry robot is mounted on top of the rack body arranged side by side along the first direction, and the three-way orthogonal spatial layout defines a dedicated and interference-free lateral movement path across the aisle, allowing the climbing robot to transfer across the rack without detours, significantly reducing the time spent on spatial transfer of equipment across the aisle; the scheduling module accurately matches the climbing robot with the task to be executed and issues instructions to realize the gantry robot's directional grasping and transfer of the target climbing robot, avoiding the ineffective operation of blind scheduling and repeated matching, and reducing the waiting time for task connection; at the same time, the gantry robot drives the climbing robot. Cross-aisle operations enable a single climbing robot to overcome the limitations of single-shelf operations and serve multiple sets of shelves, improving equipment utilization. This eliminates the need for additional dedicated shelves for cross-aisle goods retrieval and delivery, avoiding waste of equipment resources and scheduling redundancy. Furthermore, the scheduling module coordinates the collaborative operation of the gantry and climbing robot, achieving unified scheduling and execution of cross-aisle tasks. This streamlines the goods scheduling process between multiple shelves, preventing backlogs caused by independent shelf operations. It allows for efficient allocation and flow of tasks across the entire area, improving goods scheduling efficiency throughout the entire process from equipment transfer, task matching, resource utilization to overall scheduling. Attached Figure Description

[0017] Figure 1 This is a schematic diagram of a racking system for cross-aisle operations according to an embodiment of the present invention; Figure 2 This is a schematic diagram of the gripper and corresponding station of the gantry robot according to an embodiment of the present invention.

[0018] Explanation of reference numerals in the attached figures: 1-Climbing robot; 2-Shelf body; 3-Gantry robot; 4-Gripper. Detailed Implementation

[0019] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Although some embodiments of the present invention are shown in the drawings, it should be understood that the present invention can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of the present invention. It should be understood that the accompanying drawings and embodiments of the present invention are for illustrative purposes only and are not intended to limit the scope of protection of the present invention.

[0020] It should be understood that the various steps described in the method embodiments of the present invention may be performed in different orders and / or in parallel. Furthermore, the method embodiments may include additional steps and / or omit the steps shown. The scope of the present invention is not limited in this respect.

[0021] In the attached figures, the X-axis represents the first direction, i.e., the lateral movement direction; the Y-axis represents the length direction of the shelf body, i.e., the longitudinal movement direction; and the Z-axis represents the vertical direction. It should be noted that the aforementioned representations of the X, Y, and Z axes are merely for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on the invention. The Z-axis in the attached figures represents the vertical direction, i.e., the up-down position, with the positive direction of the Z-axis representing upward and the negative direction representing downward; the X-axis in the attached figures represents the horizontal direction and is specified as the front-back position, with the positive direction of the X-axis representing the front side and the negative direction representing the rear side; and the Y-axis in the attached figures represents the left-right position, with the positive direction of the Y-axis representing the left side and the negative direction representing the right side. It should also be noted that the aforementioned representations of the Z, Y, and X axes are merely for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on the invention.

[0022] The term "comprising" and its variations as used herein are open-ended, meaning "including but not limited to"; the term "based on" means "at least partially based on"; the term "one embodiment" means "at least one embodiment"; the term "another embodiment" means "at least one additional embodiment"; the term "some embodiments" means "at least some embodiments"; and the term "optionally" means "optional embodiments". Definitions of other terms will be given in the following description. It should be noted that the concepts of "first," "second," etc., mentioned in this invention are used only to distinguish different devices, modules, or units, and are not intended to limit the order of functions performed by these devices, modules, or units or their interdependencies.

[0023] It should be noted that the terms "a" and "a plurality of" used in this invention are illustrative rather than restrictive. Those skilled in the art should understand that, unless otherwise expressly indicated in the context, they should be understood as "one or more".

[0024] The names of the messages or information exchanged between the multiple devices in the embodiments of the present invention are for illustrative purposes only and are not intended to limit the scope of these messages or information.

[0025] In related technologies, traditional racking systems often employ a fixed deployment mode with dedicated robots in a single aisle (the robot's working channel corresponding to the rack itself). Each robot is confined to its own aisle to perform bin retrieval and placement operations, unable to support other areas across aisles. This mode lacks global resource coordination capabilities, and when the workload fluctuates across different aisles, it easily leads to an imbalance between busy and idle periods: some aisles experience robot overload due to concentrated orders, resulting in task queues and backlogs; while in aisles with a sudden decrease in workload, the corresponding robots remain idle for extended periods, resulting in wasted equipment resources. Furthermore, the configuration of robots serving only fixed racks not only increases initial equipment procurement costs but also adds to the burden of subsequent maintenance. These problems directly lead to inefficient resource allocation in the warehousing system, disordered task processing rhythms, and a lack of cross-aisle collaboration capabilities, ultimately severely impacting the efficiency of goods scheduling between different racks and failing to meet the core requirements of modern warehousing: "high flexibility, low cost, and rapid turnover."

[0026] To address the problems existing in the aforementioned related technologies, embodiments of the present invention provide a racking system for cross-aisle operations.

[0027] like Figure 1 As shown in the figure, an embodiment of the present invention provides a racking system for cross-aisle operations, including a climbing robot 1, a gantry robot 3, and a scheduling module. The gantry robot 3 is used to be mounted on top of multiple shelf bodies 2 arranged side by side along the first direction, the first direction being perpendicular to the length direction of the shelf body 2, and the first direction, the length direction of the shelf body 2, and the vertical direction being perpendicular to each other.

[0028] Specifically, the gantry robot 3, as the core lateral movement carrier for cross-aisle operations, is erected on the top area of ​​multiple sets of rack bodies 2. These multiple sets of rack bodies 2 are arranged in a neat and parallel layout along the first direction, forming a multi-aisle rack cluster structure. The first direction, the length direction of the rack body 2, and the vertical direction form a pairwise perpendicular spatial orthogonal relationship, with no spatial angle overlap between the three directions, forming a clear three-dimensional spatial coordinate system. This spatial layout and orientation setting defines a dedicated working movement path for the gantry robot 3 along the first direction, enabling it to achieve stable and precise lateral movement on the top of each rack body 2. This lays the spatial and motion foundation for the subsequent grasping and climbing robot 1 to complete the position transfer across the rack body 2, adapting to the cross-area operation requirements of the multi-aisle rack system.

[0029] It should be noted that the climbing robot 1 is the core execution unit for the goods operation of a single rack body 2 in the rack operation system. Its core function is to accurately pick up and deliver goods to each layer of the rack body 2 by relying on the climbing track of the rack body 2. At the same time, it serves as the goods carrier and end-effector for cross-aisle operations. In conjunction with the gripper 4 of the gantry robot 3, it realizes the flow of goods across rack bodies 2, breaks the boundaries of single rack operation, and allows a single climbing robot 1 to serve multiple rack bodies 2, which greatly improves the equipment utilization rate and the flexibility of cross-area goods scheduling. The operation is divided into two categories, both controlled by unified instructions from the scheduling module: First, independent operation within a single shelf, where the climbing robot 1 moves vertically up and down along the dedicated climbing track of its shelf body 2 to accurately reach the designated storage location and complete the grabbing, placement, and delivery of goods, thus completing the goods flow task within a single shelf; Second, collaborative operation across aisles, where the starting shelf body 2 moves from the picking position to the preset transfer station along the original longitudinal path of the aisle. After the gantry robot 3 completes the grabbing and transfer to the corresponding station of the target shelf body 2, the climbing robot 1 detaches from the gantry robot 3 and moves along the climbing track of the target shelf body 2 along the longitudinal path of the target aisle to the target location of the goods, completing the cross-aisle delivery of goods, or in reverse, completing the transfer of goods from the target shelf to the starting shelf. Throughout the operation, the climbing robot 1 will provide real-time feedback on its position, movement, and operation status to the scheduling module to ensure precise movements and seamless connection with the gantry robot 3.

[0030] The scheduling module is used to control the gantry robot 3 to grab the corresponding climbing robot 1 according to the acquired task to be executed, and to perform cross-aisle operations between the rack bodies 2 corresponding to the task to be executed.

[0031] Specifically, the scheduling module, as the core control and collaborative scheduling unit for cross-aisle operations in the racking system, acquires all pending tasks within the system in real time. These tasks include key information such as the target rack body 2 for goods retrieval and delivery, operation priority, and location coordinates. The scheduling module first matches the corresponding climbing robot 1 based on the task information (during matching, it selects suitable execution entities based on indicators such as the climbing robot 1's health, current operation status, and scheduling cost). Then, according to preset path planning, site reservation, and conflict resolution rules, it issues precise action and movement commands to the gantry robot 3, controlling the gantry robot 3 to move in the corresponding... The starting shelf body 2 completes a stable gripping of the target climbing robot 1 at the top of the station corresponding to the starting shelf body 2. Then, it drives the gripping climbing robot 1 to move precisely laterally across the aisle along the first direction of the shelf top, between the starting shelf body 2 and the target shelf body 2 corresponding to the task to be performed, until the climbing robot 1 is transferred to the designated working position of the target shelf body 2. This realizes the cross-aisle scheduling and operation connection of the climbing robot 1, ensuring the efficient and orderly execution of cross-aisle goods picking and delivery tasks. At the same time, it coordinates the movement rhythm of the gantry robot 3 and the climbing robot 1 to avoid resource and path conflicts between the equipment.

[0032] In this embodiment, the gantry robot is mounted on top of the rack body arranged side-by-side along the first direction, and the three-way orthogonal spatial layout defines a dedicated, interference-free lateral movement path across aisles for it. This allows the climbing robot to transfer across racks without detours, significantly reducing the time required for equipment to transfer across aisles. The scheduling module accurately matches the climbing robot with the task to be executed and issues instructions, enabling the gantry robot to directionally grasp and transfer the target climbing robot. This avoids ineffective operations such as blind scheduling and repeated matching, reducing the waiting time for task connection. At the same time, the gantry robot drives the climbing robot to work across aisles, enabling single... The climbing robot can break through the limitations of single-shelf operation, serving multiple sets of shelves, improving equipment utilization, and eliminating the need for additional shelves-specific robots for cross-aisle goods retrieval and delivery, thus avoiding waste of equipment resources and scheduling redundancy. In addition, the scheduling module coordinates the collaborative operation of the gantry and climbing robot, realizing unified scheduling and execution of cross-aisle tasks, opening up the goods scheduling link between multiple shelves, avoiding the backlog of goods scheduling caused by independent operation of each shelf, and enabling efficient allocation and flow of tasks to be executed throughout the entire area, improving the efficiency of goods scheduling from equipment transfer, task matching, resource utilization to overall scheduling.

[0033] Optionally, controlling the gantry robot to grasp the climbing robot according to the acquired task to be executed, and performing cross-aisle operations between the rack bodies corresponding to the task to be executed, includes: Construct a cost matrix based on all the climbing robots and the tasks to be performed; The task execution scheme with the minimum cost is obtained by the Hungarian algorithm based on the cost matrix, wherein the task execution scheme includes a one-to-one correspondence between the climbing robot and the task to be executed; According to the task execution plan, the gantry robot is controlled to cooperate with the climbing robot to perform the cross-tunnel operation.

[0034] Optionally, the method for constructing the cost matrix includes: Based on each climbing robot and the task to be performed, the comprehensive cost of the corresponding elements in the cost matrix is ​​obtained through a preset cost relationship; The cost matrix is ​​generated by combining the costs of all the elements.

[0035] Optionally, the cost relationship satisfies: C ij =w1×D ij +w2×SOC i +w3×TQ j +w4×TC ij ; Among them, C ij For the i-th climbing robot performing the j-th task to be performed, D is the comprehensive cost of the element. ij SOC is the distance from the i-th climbing robot to the j-th task point of the task to be performed. i Let TQ be the battery level of the i-th climbing robot. j TC is the waiting time for the j-th task to be executed. ij The time for the i-th climbing robot to perform the j-th task across the tunnel is given by w1, w2, w3, and w4, which are the corresponding weighting coefficients.

[0036] In this optional embodiment, based on all climbing robots and the set of tasks to be executed, including cross-lane tasks, an n×m dimension comprehensive cost matrix is ​​constructed (where n is the number of climbing robots and m is the number of tasks to be executed). Each element C in the matrix... ij Let w1 represent the comprehensive cost of the i-th climbing robot performing the j-th task. Its calculation strictly follows a pre-defined cost relationship formula, where w1, w2, w3, and w4 are weight coefficients dynamically optimized through reinforcement learning (Q-learning), and D... ij SOC represents the straight-line distance from the i-th climbing robot to the j-th target storage location (task point) for the task to be performed. i TQ represents the current remaining battery power (in percentage) of the i-th climbing robot. j TC is the cumulative waiting time of the j-th task in the queue. ijThe total time spent on lateral transfer (cross-lane transfer) when the i-th climbing robot performs the j-th task is considered (this value is 0 for the task in this lane). It should be noted that, to eliminate the unreasonable influence of differences in units and numerical scales between different features (parameters), the above data needs to be normalized before weighted summation. For example, assuming a system has 3 climbing robots and several tasks to be performed, the collected data includes: distance D... ij The unit is meters (e.g., 5m, 20m), and the remaining power SOC. i The percentage (e.g., 60%, 85%) and the waiting time (TQ) are used to indicate the time elapsed. j The unit is seconds (e.g., 30s, 150s), and the time taken to cross a tunnel is TC. ij The unit is also seconds (e.g., 0s, 40s). Due to the significant differences in the units and magnitudes of the parameters (e.g., a maximum distance of 20, but a waiting time of hundreds), direct weighting will lead to the large numerical parameters dominating the results. Therefore, each type of parameter needs to be normalized separately. For example, minimum-maximum normalization is used: all straight-line distances are scaled to [0, 1] according to the historical maximum / minimum distance in the system. Similarly, the remaining power, cumulative waiting time and cross-lane time (note that power often needs to be reverse normalized, because higher power is better) are normalized independently to eliminate the influence of dimensions and ensure that the weighted summation result reasonably reflects the overall scheduling priority. Then, the Hungarian algorithm is called to solve the overall cost matrix. The Hungarian algorithm is used to solve the assignment problem (allocation problem), that is, in the given cost matrix, a column (personnel) is assigned to each row (task) so that the total cost is minimized and each row and column is assigned only once. The algorithm reduces the number of zero elements in the matrix by row reduction and column reduction, and then covers all zeros with the fewest covering lines. If the number of covering lines is equal to the matrix order, the optimal assignment can be found directly. Otherwise, the matrix is ​​adjusted and the iteration continues until the optimal solution is found. This algorithm finds the optimal matching scheme that minimizes the total global cost by considering all robot-task pairings. The resulting task execution scheme clarifies the one-to-one correspondence between each climbing robot and a single task to be executed, ensuring that the task allocation not only meets the global optimization requirements but also achieves load balancing among multiple robots. This lays the foundation for subsequent path planning, site reservation, and conflict resolution. Compared with traditional allocation methods, it significantly shortens the average task completion time and improves equipment utilization.

[0037] Furthermore, after obtaining the task execution scheme with the lowest global cost using the Hungarian algorithm, collaborative operation instructions are issued to each device based on the one-to-one correspondence between the climbing robot and the cross-aisle task to be performed in the scheme. This coordinates the control of the gantry robot and the climbing robot to cooperate in completing the cross-aisle operation. In specific execution, the climbing robot to perform the task is first set up on the corresponding starting shelf body. The scheduling module sends a work preparation instruction to the climbing robot. After the climbing robot completes the current task and returns to the designated gripping position (station) at the top of the shelf body, the scheduling module issues movement and gripping instructions to the matching gantry robot, controlling the gantry robot to move precisely along the first direction at the top of the shelf to the designated gripping position (station). The system securely grasps the climbing robot at the grab position, then moves it along the planned transverse path across the aisle, precisely transferring it between the starting shelf and the target shelf for the task, until the climbing robot is placed at the preset work position (corresponding station) on the target shelf. After placement, the gantry robot returns to its original position or executes the next task according to instructions, while the climbing robot performs goods retrieval and delivery operations on the target shelf. The entire process is monitored in real time by the scheduling module, which monitors the movement status and progress of both devices and dynamically adjusts the action instructions according to the actual working conditions to ensure that the movement rhythm of the two devices is highly coordinated and that there are no path conflicts, thus ensuring that the cross-aisle operation is completed efficiently and orderly in strict accordance with the task execution plan.

[0038] Optionally, controlling the gantry robot to cooperate with the climbing robot to perform the cross-tunnel operation according to the task execution plan includes: Determine the corresponding starting shelf body and target shelf body based on the task to be executed; The climbing robot corresponding to the task to be performed is set on the corresponding climbing track of the starting shelf body; The gantry robot is controlled to cooperate with the corresponding climbing robot to perform the cross-aisle operation according to the cross-aisle movement path of the task to be performed. The cross-aisle movement path includes the original longitudinal path of the climbing robot from the picking position of the starting shelf body to the corresponding station, the cross-aisle transverse path of the gantry robot grabbing from the station of the starting shelf body to the station of the target shelf body, and the target longitudinal path of the climbing robot from the station of the target shelf body to the corresponding target position.

[0039] In this optional embodiment, the core parameters such as the location information and operational requirements of the goods to be performed for the cross-aisle task are used to accurately determine the starting and target rack bodies corresponding to the task, clarifying the start and end points of the cross-aisle operation. Then, based on task matching rules, the climbing robot corresponding to the task to be performed is scheduled to the dedicated climbing track of the starting rack body, ensuring that the climbing robot is in an operationally compatible state with the starting rack, preparing for subsequent goods retrieval and cross-aisle transfer. Next, the scheduling module retrieves the pre-planned cross-aisle movement path and coordinates the gantry robot and the corresponding climbing robot to perform the cross-aisle operation according to this path. This cross-aisle movement path is a three-segment integrated path, specifically including the path of the climbing robot moving from the goods retrieval position of the starting rack body along the original longitudinal path of the aisle to the preset transfer station of the rack, the path of the gantry robot grabbing the climbing robot from the starting rack station, and the precise... The system outlines the path from the target shelf to the corresponding transfer station, and the path from the target shelf station to the target location of the goods via the longitudinal path of the target aisle after the climbing robot is deployed. During operation, the climbing robot first retrieves the goods along the original longitudinal path of the aisle and moves them to the starting station. The gantry robot then uses the side camera of the gripper to scan the QR code for positioning and infrared beam calibration. After a stable grip with the climbing robot at the station, the gantry robot moves along the lateral path to complete the cross-aisle transfer and deploys the goods to the corresponding station of the target shelf. Finally, the climbing robot moves along the longitudinal path of the target aisle to the target location to complete the delivery of the goods. The entire scheduling module synchronizes the movement status and operation progress of the two devices in real time and issues collaborative action instructions according to the connection nodes of the three-segment path to ensure that the movement rhythm of the gantry robot and the climbing robot are highly matched and the operation actions are seamlessly connected, ensuring that the cross-aisle operation is completed efficiently and accurately according to the planned path.

[0040] Optionally, the system also includes: Obtain the actual arrival time of the climbing robot at the corresponding cross-lane station and the corresponding scheduled time window; If the actual arrival time is within the reservation time window, then the actual arrival time is determined as the start time of the reservation time window; Otherwise, execute the preset conflict resolution strategy.

[0041] In this optional embodiment, the actual arrival time of each climbing robot at the station corresponding to the shelf body (i.e., the stations corresponding to the starting shelf body and the target shelf body in a task to be executed) is obtained, as well as the pre-allocated reservation time window for that station (format: [T_arrive, T_arrive+T_occupy], where T_arrive is the initial planned arrival time, T_occupy is the station occupation time, and the window includes a ±30-second elastic buffer time, allowing the actual arrival time to fluctuate within ±30 seconds of the reservation time). If the actual arrival time of the climbing robot is within the reservation time window (including the elastic buffer range), the actual arrival time is automatically updated to the planned arrival time of the reservation time window to ensure that subsequent operations are connected in an orderly manner without additional scheduling adjustments. If the actual arrival time exceeds the reservation time window range, a preset conflict resolution strategy is immediately triggered to quickly resolve resource competition and avoid task backlog.

[0042] Optionally, the conflict resolution strategy includes: Obtain the available time before and after the scheduled time window; If the actual arrival time is within the range of the pre- and post-arrival free time, and the time interval between the actual arrival time and the initial planned arrival time of the reservation window is less than a preset adjustment threshold, then the actual arrival time is determined as the planned arrival time. Otherwise, the climbing robot's cross-lane operation is dynamically adjusted, wherein the dynamic adjustment includes site switching, task reallocation, and task priority preemption.

[0043] In this optional embodiment, the conflict resolution strategy is specifically implemented as follows: First, obtain the idle time before and after the corresponding reservation time window (i.e., the unoccupied time period before the reservation window start time T_arrive, and the available time period after the end time T_arrive + T_occupy); then determine whether the robot's actual arrival time is within the range of the idle time before and after, and whether the time interval between the actual arrival time and the initial planned arrival time T_arrive (the planned arrival time of the original reservation time window) is less than a preset adjustment threshold (set to 5 minutes by default). If both conditions are met, the actual arrival time is directly updated to the new planned arrival time without changing the site and task allocation, and the conflict is resolved only through a small time adjustment; if either condition is not met (such as exceeding the range of the idle time before and after, or the time interval exceeding 5 minutes), the climbing robot's cross-lane operation is dynamically adjusted: site switching is performed first, and other available cross-lane sites are queried, such as... Figure 1 and Figure 2As shown, multiple grippers 4 can be arranged side by side along the length of the rack body 2 on the gantry robot 3. Each gripper 4 can move above the rack body 2 along the first direction, that is, to grab the climbing robot 1 to work across the rack. Therefore, multiple stations (stations DZ1, DZ2, DZ3) corresponding to the grippers are generated. The score for each site is calculated using a multi-site selection scoring formula (Score_site = φ1 × (1 / D_total) + φ2 × (1 / T_wait) + φ3 × Priority_site). The system then switches to the optimal site. Here, φ1, φ2, and φ3 are preset weighting coefficients (ranging from 0 to 1, summing to 1, used to adjust the influence of each factor on the score); D_total refers to the total path length (in meters) of the climbing robot across the alleyway at that site, calculated as the sum of the distance from the current location to the site, the lateral transfer distance, and the distance from the site to the target location, with the reciprocal (1 / D_total) used in the formula, indicating that a shorter path corresponds to a higher score; T_wait refers to the average waiting time (in seconds) of the current reservation queue at that site, used to determine the next site based on the site reservation table. The score is determined by taking the reciprocal (1 / T_wait) of the idle time window. The shorter the waiting time, the higher the score. Priority_site refers to the overall priority score of the site (ranging from 0 to 1), which is calculated by "health evaluation value × 1.2 - number of failures × 0.1". It directly participates in the scoring. The higher the site priority, the higher the corresponding score. D_total, T_wait and Priority_site are also normalized to eliminate differences in units. If site switching is not feasible, the Hungarian algorithm is run again to redistribute the task, assigning the task to other climbing robots with qualified health evaluation values ​​and no conflicts. If the first two methods still cannot resolve the conflict, priority preemption and task execution priority ranking can be determined based on the priority of the current pending task and the priority of the conflicting pending task. For example, if both the currently pending task and the conflicting pending task correspond to station DZ1, and the priority of the currently pending task is ≥8 while the priority of the conflicting pending task is <5, the task priority preemption mechanism is triggered, forcibly occupying the reserved resources of the lower priority task. The preempted task is automatically postponed or rerouted (i.e., based on real-time device status, station occupancy, remaining task queue, and other data, a complete execution path and work plan are re-planned for it). At the same time, the number of preemptions is recorded to prevent a robot from being preempted for a long time, ensuring that conflict resolution is efficient and does not affect the overall work rhythm of the system.

[0044] Optionally, the system also includes: Obtain the status data of the climbing robot corresponding to the task to be executed; If the status data meets the feasibility conditions for crossing the tunnel, then it is determined that the climbing robot can perform the tunnel crossing operation; Otherwise, select the climbing robot again.

[0045] In this optional embodiment, the status data of each initial climbing robot is obtained, including but not limited to its current position, battery level, current task status (idle / running / faulty), mobility and the lane it is in. Based on the above information, it is determined whether the robot meets the conditions for crossing lanes (e.g., the robot is in an idle or interruptible state, the remaining battery is sufficient to complete the movement and operation across lanes, the mechanical structure is normal, etc.). When the conditions are met, the robot is used as a climbing robot to perform the crossing lane task, that is, it is qualified to participate in the allocation of crossing lane tasks.

[0046] Optionally, the status data includes health data and task data, wherein the health data includes corresponding battery health factors, mechanical health factors, communication health factors, and historical health factors; the step of determining that the climbing robot can perform the cross-lane operation when the status data meets the cross-lane feasibility conditions includes: The initial health evaluation value of the climbing robot is obtained by multiplying the battery health factor, the mechanical health factor, the communication health factor, and the historical health factor. When the health evaluation value is greater than the preset health threshold, the climbing robot will meet the schedulable requirements. If the task data corresponding to the climbing robot that meets the schedulable requirements satisfies the cross-lane feasibility condition, then it is determined that the climbing robot can perform the cross-lane operation.

[0047] In this optional embodiment, the status data of the climbing robot corresponding to each task to be executed is comprehensively analyzed to accurately assess the robot's operational capabilities. This status data includes two parts: the robot's health data and the task data related to the task. The health data is composed of battery health factors (reflecting remaining battery power and battery degradation), mechanical health factors (characterizing the integrity of mechanical components such as drive, lifting, and gripping), communication health factors (measuring the stability and latency of communication with the scheduling system), and historical health factors (derived based on statistics such as historical failure rates, task completion success rates, and maintenance records). The scheduling module first multiplies the above four health factors to calculate the robot's health evaluation value. This product form reflects the design principle that the system is highly sensitive to degradation of any critical subsystem. When the health evaluation value is greater than a preset health threshold, the climbing robot is determined to be in a reliable operating state and is identified as a schedulable robot. For example, the battery health factor is calculated from the climbing robot's battery charge / discharge cycle count, capacity degradation rate, and charge / discharge stability. Specifically, since the initial nominal battery capacity is C0 and the current capacity degradation rate is 6%, the current remaining capacity is C = C0 × (1 / 2) * ... 6%) = 0.94C0, the number of charge-discharge cycles (320 times) is used to verify whether the degradation is within a reasonable range. The health factor is mainly determined by the capacity retention rate. The charge-discharge stability is good (no abnormal fluctuations), so no additional penalty term is introduced. Therefore, the battery health factor is defined as the ratio of the current available capacity to the initial capacity, combined with stability fine-tuning: H_battery = (1 Capacity decay rate) × α = , where α is the stability correction coefficient (taken as 0.99~1.0 when stable). Substituting the data, we get H_battery = (1 0.06)×0.989≈0.93; Based on mechanical operation data such as the wear degree of the climbing mechanism, the flexibility of the lateral movement joints, and the gripper's grasping accuracy, combined with recent records of no mechanical failures, a mechanical health factor is derived. Specifically, three key mechanical performance indicators and their current evaluation values ​​are set (all normalized to [0, 1], with higher values ​​indicating better condition): Climbing mechanism wear degree: the less wear, the better, with a corresponding health level of 0.94 after testing; Lateral movement joint flexibility: the measured motion response is smooth, with a health level of 0.97; Gripper grasping accuracy: the positioning error is within the allowable range, with a health level of 0.98. According to the influence of each component on the overall mechanical performance, weights are assigned (if wear is more critical, the weight is slightly higher): wear degree weight w1=0.4, joint flexibility weight w2=0.35, grasping accuracy weight w3=0.25. At the same time, since the system has no recent mechanical failure records, a reliability enhancement factor β=1.0 is introduced (β<1 if there is a failure). The mechanical health factor is: H_mechanical=β×(w1×0.94+w2×0.97+w3×0.98)≈0.96, therefore, H_mechanical=0.96. Based on the average communication delay, signal stability, and packet loss rate between the climbing robot and the scheduling module, and the cross-aisle shelf components, the communication health factor is determined. Specifically, each communication performance indicator is first normalized to the [0,1] interval (the larger the value, the better the communication status): Average communication delay: The measured value is 10ms. Assuming the maximum allowable delay of the system is 50ms, the delay health is hdelay=1. 10 / 50 = 0.80, Signal stability: evaluated by signal-to-noise ratio or connection fluctuation, assuming the current rating hstability is 0.95 (high stability); Packet loss rate: measured at 0.2%. Assuming a tolerance limit of 2%, the packet loss health is hloss = 1. 0.2% / 2%=1 0.1 = 0.90, setting weights to reflect the importance of each factor (e.g., delay and packet loss are more critical for real-time control): delay weight w1 = 0.4, signal stability weight w2 = 0.3, packet loss rate weight w3 = 0.3, then the communication health factor H_communication = w1 × hdelay + w2 × hstability + w3 × hloss = 0.4 × 0.80 + 0.3 × 0.95 + 0.3 × 0.90 ≈ 0.88. Therefore, the communication health factor H_communication = 0.88 is determined. By analyzing the job failure rate, task completion quality, and fault recovery efficiency over the past 30 days, the historical health factor is obtained. Specifically, the three historical operating indicators are normalized to the [0, 1] interval (the higher the value, the better the condition): job failure rate: 1.2% over the past 30 days. Assuming an acceptable upper limit of 5%, the lower the failure rate, the better, and its health level is hfault = 1. 1.2% / 5%=1 0.24 = 0.76, Task completion quality: Evaluated based on task success rate, bin placement accuracy, etc., assuming a comprehensive score of 96%, i.e., hquality = 0.96. Fault recovery efficiency: Measured by average fault recovery time. If the current recovery speed is better than the benchmark (e.g., 90% of faults are recovered within 2 minutes), then the health factor hrecovery = 0.92 is set. Weights are assigned to reflect the importance of each indicator (e.g., quality is the most critical, followed by fault rate): Fault rate weight w1 = 0.4, task quality weight w2 = 0.4, recovery efficiency weight w3 = 0.2, then the historical health factor Hhist The formula is: orical = w1 × hfault + w2 × hquality + w3 × hrecovery = 0.4 × 0.76 + 0.4 × 0.96 + 0.2 × 0.92 ≈ 0.87. Therefore, the historical health factor Hhistorical = 0.87. The final health evaluation value of the robot is 0.93 × 0.96 × 0.88 × 0.87 ≈ 0.684. Since it is higher than the preset health threshold of 0.6, it is determined that the schedulable robot can participate in cross-lane operations normally. If it is between 0.5 and 0.6, it is prohibited from participating in cross-lane operations. If it is less than 0.5, it is suspended from use.

[0048] Furthermore, by combining the task data corresponding to the schedulable robot (such as current task status, location, whether it is idle, and whether it is in an interruptible phase), it is determined whether it meets the preset conditions for cross-lane feasibility. Only when the task data meets these conditions is the climbing robot finally confirmed as a climbing robot capable of performing the corresponding task and included in the subsequent task allocation process. Through this dual screening mechanism of "health assessment + task status," robots that are truly suitable for performing cross-lane operations are dynamically and accurately identified while ensuring operational safety and equipment reliability, providing a high-quality resource pool for achieving efficient and robust global scheduling.

[0049] Optionally, the task data includes cross-tunnel operation waiting time, cross-tunnel movement time, cross-tunnel energy consumption cost, cross-tunnel resource conflict probability, cross-tunnel benefit, task execution time in this tunnel, operation energy consumption in this tunnel, resource conflict probability in this tunnel, and task waiting time in this tunnel; the step of determining that the climbing robot can perform the cross-tunnel operation when the task data corresponding to the climbing robot that meets the schedulable requirements satisfies the cross-tunnel feasibility condition includes: Based on the cross-lane operation waiting time, the cross-lane movement time, the cross-lane energy consumption cost, the cross-lane resource conflict probability, and the cross-lane benefit, the corresponding cross-lane scheduling cost is obtained by weighted summation. Based on the task execution time of this roadway, the operation energy consumption of this roadway, the resource conflict probability of this roadway, and the task waiting time of this roadway, the corresponding scheduling cost of this roadway is obtained by weighted summation; If the product of the cross-lane scheduling cost and the preset cross-lane coefficient is greater than the current lane scheduling cost, then the climbing robot is determined to be capable of performing the cross-lane operation.

[0050] In this optional embodiment, the task data covers multi-dimensional key indicators for cross-aisle (from the starting shelf body to the target shelf body) operations, specifically including cross-aisle operation waiting time, cross-aisle movement time, cross-aisle energy consumption cost, cross-aisle resource conflict probability, cross-aisle benefits, as well as the task execution time, energy consumption, resource conflict probability, and waiting time for the current aisle operation. When determining whether the climbing robot can perform cross-aisle operations, it must be based on the above task data and follow the cross-aisle feasibility conditions: First, for cross-aisle operation scenarios, the system uses preset weight coefficients α1, α2, α3, α4, and β (corresponding to cross-aisle operation waiting time, cross-aisle movement time, cross-aisle energy consumption cost, cross-aisle resource conflict probability, and cross-aisle benefits, respectively). Before weighted summation, these parameters also need to be normalized to eliminate weight distortion and evaluation bias caused by differences in units and numerical scales, ensuring that all factors are considered correctly. In comprehensive decision-making, the system contributes fairly according to its actual importance to improve the rationality and optimization effect of cross-lane scheduling strategies. The cross-lane scheduling cost of schedulable robots is calculated by weighted summation: α1 × cross-lane operation waiting time + α2 × cross-lane movement time + α3 × cross-lane energy consumption cost + α4 × cross-lane resource conflict probability - β × cross-lane benefit (where the cross-lane benefit is given a negative weight to reflect the positive impact of benefit on cross-lane decision-making in cost evaluation). Secondly, for the specific lane operation scenario, the system uses preset weight coefficients β1, β2, β3, and β4 for the specific lane. The scheduling cost of this roadway is obtained by weighted summing of the task execution time, energy consumption, resource conflict probability, and task waiting time in this roadway. The cost is calculated as: β1 × task execution time + β2 × energy consumption + resource conflict probability + β4 × task waiting time. Before weighted summing, the task execution time, energy consumption, resource conflict probability, and task waiting time in this roadway need to be normalized separately to eliminate differences in dimensions (units) and numerical scales among these parameters. To address the weight distortion and evaluation bias caused by differences in degree, this method ensures that each factor contributes fairly to the overall decision-making process according to its actual importance, thereby improving the rationality and optimization effect of cross-lane scheduling strategies. Finally, the cross-lane scheduling cost is multiplied by the preset cross-lane coefficient. If the result is greater than the current lane scheduling cost, it indicates that the overall cost-effectiveness of cross-lane operations is better than current lane operations, thus meeting the cross-lane feasibility conditions. Consequently, the corresponding schedulable robot is identified as a climbing robot qualified for cross-lane operations, ensuring the scientific and rational nature of cross-lane decisions, avoiding ineffective cross-lane operations, and guaranteeing overall system optimization.

[0051] While the present invention has been disclosed above, its scope of protection is not limited thereto. Those skilled in the art can make various changes and modifications without departing from the spirit and scope of the present invention, and all such changes and modifications will fall within the scope of protection of the present invention.

Claims

1. A racking system for cross-aisle operations, characterized in that, Includes climbing robots, gantry robots, and scheduling modules. The gantry robot is used to be mounted on top of multiple rack bodies arranged side by side along a first direction, wherein the first direction, the length direction of the rack body, and the vertical direction are perpendicular to each other. The scheduling module is used to control the gantry robot to grab the corresponding climbing robot according to the acquired task to be executed, and to perform cross-aisle operations between the rack bodies corresponding to the task to be executed.

2. The racking system for cross-aisle operations according to claim 1, characterized in that, The step of controlling the gantry robot to grasp the climbing robot according to the acquired task to be executed, and performing cross-aisle operations between the rack bodies corresponding to the task to be executed, includes: Construct a cost matrix based on all the climbing robots and the tasks to be performed; The task execution scheme with the minimum cost is obtained by using the Hungarian algorithm based on the cost matrix, wherein the task execution scheme includes a one-to-one correspondence between the climbing robot and the task to be executed; According to the task execution plan, the gantry robot is controlled to cooperate with the climbing robot to perform the cross-tunnel operation.

3. The racking system for cross-aisle operations according to claim 2, characterized in that, The method for constructing the cost matrix includes: Based on each climbing robot and the task to be performed, the comprehensive cost of the corresponding elements in the cost matrix is ​​obtained through a preset cost relationship; The cost matrix is ​​generated by combining the costs of all the elements.

4. The racking system for cross-aisle operations according to claim 3, characterized in that, The cost relationship satisfies: C ij =w1×D ij +w2×SOC i +w3×TQ j +w4×TC ij ; Among them, C ij For the i-th climbing robot performing the j-th task to be performed, D is the comprehensive cost of the element. ij SOC is the distance from the i-th climbing robot to the j-th task point of the task to be performed. i Let TQ be the battery level of the i-th climbing robot. j TC is the waiting time for the j-th task to be executed. ij The time for the i-th climbing robot to perform the j-th task across the tunnel is given by w1, w2, w3, and w4, which are the corresponding weighting coefficients.

5. The racking system for cross-aisle operations according to claim 2, characterized in that, The step of controlling the gantry robot to cooperate with the climbing robot to perform the cross-tunnel operation according to the task execution plan includes: Determine the corresponding starting shelf body and target shelf body based on the task to be executed; The climbing robot corresponding to the task to be performed is set on the corresponding climbing track of the starting shelf body; The gantry robot is controlled to cooperate with the corresponding climbing robot to perform the cross-aisle operation according to the cross-aisle movement path of the task to be performed. The cross-aisle movement path includes the original longitudinal path of the climbing robot from the picking position of the starting shelf body to the corresponding station, the cross-aisle transverse path of the gantry robot grabbing from the station of the starting shelf body to the station of the target shelf body, and the target longitudinal path of the climbing robot from the station of the target shelf body to the corresponding target position.

6. The racking system for cross-aisle operations according to claim 5, characterized in that, Also includes: Obtain the actual arrival time of the climbing robot at the corresponding station and the corresponding reservation time window; If the actual arrival time is within the reservation time window, then the actual arrival time is determined as the planned arrival time of the reservation time window; Otherwise, execute the preset conflict resolution strategy.

7. The racking system for cross-aisle operations according to claim 6, characterized in that, The conflict resolution strategy includes: Obtain the available time before and after the scheduled time window; If the actual arrival time is within the range of the pre- and post-arrival free time, and the time interval between the actual arrival time and the initial planned arrival time of the reservation window is less than a preset adjustment threshold, then the actual arrival time is determined as the planned arrival time. Otherwise, the climbing robot's cross-lane operation is dynamically adjusted, wherein the dynamic adjustment includes site switching, task reallocation, and task priority preemption.

8. The racking system for cross-aisle operations according to claim 2, characterized in that, Also includes: Obtain the status data of the climbing robot corresponding to the task to be executed; If the status data meets the feasibility conditions for crossing the tunnel, then it is determined that the climbing robot can perform the tunnel crossing operation; Otherwise, select the climbing robot again.

9. The racking system for cross-aisle operations according to claim 8, characterized in that, The status data includes health data and task data. The health data includes corresponding battery health factors, mechanical health factors, communication health factors, and historical health factors. When the status data meets the feasibility conditions for crossing tunnels, it is determined that the climbing robot can perform the tunnel crossing operation, including: The health evaluation value of the climbing robot is obtained by multiplying the battery health factor, the mechanical health factor, the communication health factor, and the historical health factor. When the health evaluation value is greater than the preset health threshold, the climbing robot will meet the schedulable requirements. If the task data corresponding to the climbing robot that meets the schedulable requirements satisfies the cross-lane feasibility condition, then it is determined that the climbing robot can perform the cross-lane operation.

10. The racking system for cross-aisle operations according to claim 9, characterized in that, The task data includes cross-tunnel operation waiting time, cross-tunnel movement time, cross-tunnel energy consumption cost, cross-tunnel resource conflict probability, cross-tunnel benefit, task execution time in this tunnel, operation energy consumption in this tunnel, resource conflict probability in this tunnel, and task waiting time in this tunnel; when the task data corresponding to the climbing robot that meets the schedulable requirements satisfies the cross-tunnel feasibility condition, it is determined that the climbing robot can perform the cross-tunnel operation, including: Based on the cross-lane operation waiting time, the cross-lane movement time, the cross-lane energy consumption cost, the cross-lane resource conflict probability, and the cross-lane benefit, the corresponding cross-lane scheduling cost is obtained by weighted summation. Based on the task execution time of this roadway, the operation energy consumption of this roadway, the resource conflict probability of this roadway, and the task waiting time of this roadway, the corresponding scheduling cost of this roadway is obtained by weighted summation; If the product of the cross-lane scheduling cost and the preset cross-lane coefficient is greater than the current lane scheduling cost, then the climbing robot is determined to be capable of performing the cross-lane operation.

Images