An automated warehouse logistics management system

Abstract

This invention belongs to the field of warehousing and logistics management technology, specifically referring to an automated warehousing and logistics management system. The system includes a palletizing generation module, a refined scenario verification module, and a back-end logistics adaptive decision-making module. This solution utilizes a lightweight model without graphical rendering to verify the accessibility and collision-free nature of goods placement in real time. Through both lightweight and refined verification, backup candidate goods reduce the number of backtracking adjustments, and the optimal solution is output after iterative convergence, improving palletizing optimization efficiency and solution adaptability. The back-end logistics adaptive decision-making module constructs a dataset for the optimal path algorithm of the traveling salesman problem, collects warehouse layout data to classify layout types, gathers logistics requirements to obtain constraints, assigns weights according to layout matching and constraint matching, and automatically searches for suitable picking and scheduling strategies. This achieves precise matching between warehouse layout and logistics requirements, improving the efficiency and rationality of back-end logistics picking and scheduling.

Description

Technical Field

[0001] This invention relates to the field of warehousing and logistics management technology, specifically to an automated warehousing and logistics management system. Background Technology

[0002] The warehouse logistics automation management system is a management system that uses deep learning technology to optimize warehouse palletizing schemes and back-end logistics strategies.

[0003] Among the existing approximate solutions, such as CN119761932B, a warehouse logistics management method based on logistics big data, this solution addresses the technical problems of traditional logistics management, such as difficulty in handling multi-source heterogeneous data, lack of flexibility in inventory management, poor supply chain coordination, and especially low efficiency during e-commerce promotions and sudden public events. By collecting multi-source heterogeneous data such as logistics transportation, orders, warehousing, weather, traffic, and equipment, and using an improved automata algorithm to construct grid space mapping data, it presets four types of state transition rules for vehicles, obstacles, weather, and traffic. After iteratively updating the grid state, it extracts clustering, dynamic, and topological features to divide into M types of logistics route modes. Based on the M modes, it constructs a multi-objective optimization function and uses a molecular swarm optimization algorithm to calculate the optimal logistics planning scheme. This achieves the technical effect of accurately capturing the dynamic features of logistics data and improving risk resistance. However, it still has technical problems such as the traditional palletizing scheme generation ignoring the actual operational constraints of warehouse robot accessibility and collision, low iterative optimization efficiency, and lack of guarantee of scheme feasibility. Furthermore, for example, CN120181761B, a logistics and warehousing management system based on big data, addresses the technical problems of traditional warehousing management that fail to consider the impact of task distribution and randomness on the efficiency of transport equipment, have a single vehicle scheduling method, and are unable to meet the long-term task requirements of complex scenarios, resulting in low overall operating efficiency and numerous route conflicts. By determining effective task points based on task update frequency, selecting uniformly divided areas based on the number and distribution coefficient of effective task points, determining the scheduling method based on the task dispersion and the number of responsive vehicles in two types of areas, and determining the scheduling range based on task fluctuation coefficient and obstacle complexity, the technical effect of making area division and vehicle classification more in line with actual task distribution and optimizing warehousing management efficiency still exists. However, there are still technical problems such as diverse warehouse layout types, large differences in the arrangement of wide and narrow aisles, and varying logistics demand constraints, resulting in low efficiency and insufficient accuracy in the adaptation of picking and scheduling strategies. Summary of the Invention

[0004] To address the aforementioned issues and overcome the shortcomings of existing technologies, this invention provides an automated warehouse logistics management system. Addressing the problems of traditional palletizing scheme generation neglecting practical operational constraints such as accessibility and collision prevention for warehouse robots, and exhibiting low iterative optimization efficiency and lack of guaranteed feasibility, this solution employs a warehouse palletizing generation module using a genetic algorithm combined with a lightweight digital model. Candidate palletizing schemes are screened through selection, crossover, and mutation operations. The lightweight model, devoid of graphical rendering, verifies the accessibility and collision-free nature of goods placement in real time. Through this dual verification of lightweight and refined processes, backup candidate goods reduce the number of backtracking adjustments. After iterative convergence, the optimal scheme is output, dynamic process risks are identified and feasibility is marked, achieving end-to-end palletizing risk management driven by a digital model. Control measures ensure the solution adapts to the physical scenario, providing precise quantitative data support for implementation and improving palletizing optimization efficiency and solution adaptability. Addressing the technical challenges of diverse warehouse layouts, significant differences in aisle width and width, and varying logistics constraints, which lead to low efficiency and insufficient accuracy in picking and scheduling strategies, this solution constructs an optimal path algorithm dataset for the Traveling Salesman Problem through a backend logistics adaptive decision-making module. It standardizes the definition of aisle width and width, collects warehouse layout data to classify layout types, gathers logistics requirements to obtain constraints, and uses multi-dimensional weighted cosine similarity technology to retrieve the dataset. Weights are assigned based on layout matching and constraint matching to automatically search for suitable picking and scheduling strategies, achieving precise adaptation between warehouse layout and logistics needs, and improving the efficiency and rationality of backend logistics picking and scheduling.

[0005] The technical solution adopted by the present invention is as follows: The present invention provides an automated warehouse logistics management system, which includes a warehouse palletizing generation module, a refined scenario verification module, and a back-end logistics adaptive decision-making module; The warehouse palletizing generation module adopts a warehouse palletizing generation method and uses a genetic algorithm to generate a goods placement scheme for palletizing, thereby realizing digital model-driven end-to-end palletizing risk control and conducting preliminary verification of the goods placement scheme. The refined scenario verification module adopts a refined scenario verification method to screen goods placement schemes from a global perspective and perform full-process verification of the final state of palletizing, thereby achieving dual verification of goods placement. The back-end logistics adaptive decision-making module uses a back-end logistics adaptive decision-making method to classify the current warehouse layout and automatically search for picking scheduling strategies that are suitable for the current warehouse picking needs.

[0006] Furthermore, the warehouse palletizing generation method specifically includes the following steps: Step A1: Parameter initialization configuration, specifically, input... The system includes high-precision digital models of 3D warehouse robots and cargo, records stacking rules for each type of cargo, sets operational space margins based on the model of the warehouse robot and the size of the end effector, and clarifies optimization goals. Step A2: Initial population generation, randomizing cargo placement schemes; Step A3: Iterative optimization and lightweight verification, used for dynamic filtering during the optimization process. Specifically, prioritizing the basic principles of heavy-to-light and large-to-small, candidate sequences for cargo placement are selected using a genetic algorithm through selection, crossover, and mutation operations. This simplifies the high-precision digital model of the 3D warehouse robot, obtaining a lightweight digital model that retains only the core data of the warehouse robot without graphical rendering. During the iterative process of the genetic algorithm, it verifies in real time whether the coordinates of the cargo in the placement scheme are within the workspace of the warehouse robot, and whether the placement posture causes interference between the warehouse robot body and the cargo. The core data includes joint torque limits, reachable radius, and maximum load mass of the end effector. Each position in the candidate sequence includes backup candidate cargo, which is used to reduce the number of iteration backtracking. Each cargo placement scheme is verified by the lightweight digital model. If a cargo meets the accessibility and collision-free criteria, its placement and orientation are immediately locked, and the process proceeds to recursively verifying the next cargo. If a cargo does not meet the accessibility and collision-free criteria, the next cargo is adjusted. After verifying the position and orientation of the goods in the order of priority, check again. If the result of the second check is still not satisfactory, switch to the backup candidate goods for re-checking. If the first When all positions and orientations of all alternative candidate goods in the order of priority cannot meet the requirements, i.e., the current order sequence is infeasible, the recursive algorithm automatically backtracks to the next step. Priority goods, readjusted The order of placement and orientation of goods is determined, then the current item is returned to its position. The goods in sequence are re-verified until the next item. A feasible solution is found by sequence, where the solution represents the locked first... The placement and orientation of the goods are ordered, and the solution that maximizes the optimization objective is retained. The optimal solution is then selected for the next round of selection. This process is repeated iteratively until the optimization objective converges. Step A4: Optimal solution output, specifically, outputting the clear order of goods placement, the three-dimensional coordinates and orientation of each goods placement, forming a palletizing solution that can be directly executed.

[0007] Furthermore, the refined scene verification method specifically includes the following steps: Step B1: Set up the simulation environment, including the palletizing scheme, The high-precision digital model of the 3D warehouse robot and the digital model of the goods are imported into the simulation engineering software to restore the real warehouse storage scene, including the location of obstacles. Step B2: Refined scenario verification, used for simulation operation and real-time verification. According to the order, position and posture of goods placement in the palletizing scheme, the complete actions of picking, moving and placing that cannot be covered by the lightweight digital model are gradually simulated by simulating the palletizing process of the warehouse robot. Dynamic process risks are identified, the feasibility of the palletizing scheme is marked, and the physical scenario is implemented. Step B3: Quantitative index output. After the simulation is completed, the simulation engineering software automatically counts and records key indicators, which consist of space utilization and palletizing completion time. Step B4: Iterative correction of the solution. If the palletizing solution is marked as infeasible, return to step A3 of the warehouse palletizing generation module for re-optimization. If it is feasible and the key indicators meet the requirements, the final palletizing solution is output.

[0008] Furthermore, the adaptive decision-making method for backend logistics specifically includes the following steps: Step C1: Construct an algorithm dataset for solving the optimal path based on the Traveling Salesman Problem; Step C2: Predefined, specifically, the wide aisles represent aisles that do not directly store goods, but are only used for the movement and path switching of warehouse robots. They are few in number and wide enough to accommodate at least two warehouse robots. The narrow aisles represent the core areas directly associated with goods storage and picking. They are numerous and narrow, and can only accommodate one warehouse robot. Step C3: Define the current warehouse layout, collect the arrangement of wide and narrow aisles in the current warehouse, classify the current warehouse layout to obtain the warehouse layout type, collect the current warehouse logistics requirements, obtain the traveling salesman problem (TSP) requirement constraints, and retrieve the picking and scheduling strategy suitable for the current warehouse from the algorithm dataset based on the warehouse layout type and the TSP requirement constraints.

[0009] The beneficial effects achieved by the present invention using the above solution are as follows: (1) In response to the technical problems of traditional palletizing scheme generation ignoring the actual operational constraints such as the accessibility and collision of warehouse robots, and the low efficiency of iterative optimization and the lack of guarantee of scheme feasibility, this scheme adopts a genetic algorithm combined with a lightweight digital model through the warehouse palletizing generation module. Candidate palletizing schemes are screened through selection, crossover and mutation operations. The accessibility and collision-free nature of goods placement are verified in real time using a lightweight model without graphics rendering. In the dual verification of lightweight and fine, the number of backtracking adjustments is reduced by using backup candidate goods. After iterative convergence, the optimal scheme is output, dynamic process risks are investigated and feasibility is marked, realizing full-link palletizing risk control driven by digital model, ensuring that the scheme is adapted to the physical scenario, providing accurate quantitative data support for implementation, and improving the efficiency of palletizing optimization and the adaptability of the scheme.

[0010] (2) In response to the technical problems of low efficiency and insufficient accuracy of picking and scheduling strategies due to the diverse types of warehouse layouts, large differences in the arrangement of wide and narrow aisles, and different constraints of logistics demand, this solution constructs the optimal path algorithm dataset for the traveling salesman problem through the back-end logistics adaptive decision module, standardizes the definition of wide and narrow aisles, collects warehouse layout data to classify layout types, collects logistics demand to obtain constraints, uses multi-dimensional weighted cosine similarity technology to retrieve the dataset, assigns weights according to layout matching and constraint matching, and automatically searches for suitable picking and scheduling strategies to achieve accurate matching between warehouse layout and logistics demand, thereby improving the efficiency and rationality of back-end logistics picking and scheduling. Attached Figure Description

[0011] Figure 1 The present invention provides a module connection diagram for an automated warehouse logistics management system.

[0012] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used together with the embodiments of the invention to explain the invention and do not constitute a limitation thereof. Detailed Implementation

[0013] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0014] Example 1: See Figure 1 This embodiment provides an automated warehouse logistics management system, which includes a warehouse palletizing generation module, a refined scenario verification module, and a back-end logistics adaptive decision-making module. The warehouse palletizing generation module adopts a warehouse palletizing generation method and uses a genetic algorithm to generate a goods placement scheme for palletizing, thereby realizing digital model-driven end-to-end palletizing risk control and conducting preliminary verification of the goods placement scheme. The refined scenario verification module adopts a refined scenario verification method to screen goods placement schemes from a global perspective and perform full-process verification of the final state of palletizing, thereby achieving dual verification of goods placement. The back-end logistics adaptive decision-making module uses a back-end logistics adaptive decision-making method to classify the current warehouse layout and automatically search for picking scheduling strategies that are suitable for the current warehouse picking needs.

[0015] Example 2: See Figure 1 This embodiment is based on the above embodiment, and the warehouse palletizing generation method specifically includes the following steps: Step A1: Parameter initialization configuration, specifically, input... The system includes high-precision digital models of 3D warehouse robots and cargo, records stacking rules for each type of cargo, sets operational space margins based on the model of the warehouse robot and the size of the end effector, and clarifies optimization goals. Step A2: Initial population generation, randomizing cargo placement schemes; Step A3: Iterative optimization and lightweight verification, used for dynamic filtering during the optimization process. Specifically, prioritizing the basic principles of heavy-to-light and large-to-small, candidate sequences for cargo placement are selected using a genetic algorithm through selection, crossover, and mutation operations. This simplifies the high-precision digital model of the 3D warehouse robot, obtaining a lightweight digital model that retains only the core data of the warehouse robot without graphical rendering. During the iterative process of the genetic algorithm, it verifies in real time whether the coordinates of the cargo in the placement scheme are within the workspace of the warehouse robot, and whether the placement posture causes interference between the warehouse robot body and the cargo. The core data includes joint torque limits, reachable radius, and maximum load mass of the end effector. Each position in the candidate sequence includes backup candidate cargo, which is used to reduce the number of iteration backtracking. Each cargo placement scheme is verified by the lightweight digital model. If a cargo meets the accessibility and collision-free criteria, its placement and orientation are immediately locked, and the process proceeds to recursively verifying the next cargo. If a cargo does not meet the accessibility and collision-free criteria, the next cargo is adjusted. After verifying the position and orientation of the goods in the order of priority, check again. If the result of the second check is still not satisfactory, switch to the backup candidate goods for re-checking. If the first When all positions and orientations of all alternative candidate goods in the order of priority cannot meet the requirements, i.e., the current order sequence is infeasible, the recursive algorithm automatically backtracks to the next step. Priority goods, readjusted The order of placement and orientation of goods is determined, then the current item is returned to its position. The goods in sequence are re-verified until the next item. A feasible solution is found by sequence, where the solution represents the locked first... The placement and orientation of the goods are ordered, and the solution that maximizes the optimization objective is retained. The optimal solution is then selected for the next round of selection. This process is repeated iteratively until the optimization objective converges. Step A4: Optimal solution output, specifically, outputting the clear order of goods placement, the three-dimensional coordinates and orientation of each goods placement, forming a palletizing solution that can be directly executed.

[0016] Example 3: See Figure 1 This embodiment is based on the above embodiment, and the refined scene verification method specifically includes the following steps: Step B1: Set up the simulation environment, including the palletizing scheme, The high-precision digital model of the 3D warehouse robot and the digital model of the goods are imported into the simulation engineering software to restore the real warehouse storage scene, including the location of obstacles. Step B2: Refined scenario verification, used for simulation operation and real-time verification. According to the order, position and posture of goods placement in the palletizing scheme, the complete actions of picking, moving and placing that cannot be covered by the lightweight digital model are gradually simulated by simulating the palletizing process of the warehouse robot. Dynamic process risks are identified, the feasibility of the palletizing scheme is marked, and the physical scenario is implemented. Step B3: Quantitative index output. After the simulation is completed, the simulation engineering software automatically counts and records key indicators, which consist of space utilization and palletizing completion time. Step B4: Iterative correction of the solution. If the palletizing solution is marked as infeasible, return to step A3 of the warehouse palletizing generation module for re-optimization. If it is feasible and the key indicators meet the requirements, the final palletizing solution is output.

[0017] Example 4: See Figure 1 This embodiment is based on the above embodiment, and the back-end logistics adaptive decision-making method specifically includes the following steps: Step C1: Construct an algorithm dataset for solving the optimal path based on the Traveling Salesman Problem; Step C2: Predefined, specifically, the wide aisles represent aisles that do not directly store goods, but are only used for the movement and path switching of warehouse robots. They are few in number and wide enough to accommodate at least two warehouse robots. The narrow aisles represent the core areas directly associated with goods storage and picking. They are numerous and narrow, and can only accommodate one warehouse robot. Step C3: Define the current warehouse layout, collect the arrangement of wide and narrow aisles in the current warehouse, classify the current warehouse layout to obtain the warehouse layout type, collect the current warehouse logistics requirements, obtain the traveling salesman problem (TSP) requirement constraints, and retrieve the picking and scheduling strategy suitable for the current warehouse from the algorithm dataset based on the warehouse layout type and the TSP requirement constraints.

[0018] Example 5: See Figure 1 This embodiment is based on the above embodiment. In step A1, the digital model of the cargo includes size, weight, and stacking constraints.

[0019] Example 6: See Figure 1 This embodiment is based on the above embodiment. In step A1, the dimensions of the end effector include the gripper width and height.

[0020] Example 7: See Figure 1 This embodiment is based on the above embodiment. In step A1, the optimization objective is specifically to maximize space utilization and minimize palletizing completion time.

[0021] Example 8: See Figure 1 This embodiment is based on the above embodiment. In step A3, the lightweight digital model only includes the reachable range and joint limits of the end effector of the warehouse robot.

[0022] Example 9: See Figure 1 This embodiment is based on the above embodiment. In step A3, the accessibility and collision-free determination criteria are used to ensure that the current goods do not overlap in size with the already placed goods and do not collide, and meet the operating space threshold requirements of the warehouse robot. The specific operation is as follows: Step A31: Global collision check. Traverse the initially formed placement sequence and check whether there is any size overlap between all goods to avoid missing local collisions during recursive layer-by-layer verification. Step A32: Global review of the operating space. Check whether the joints of the warehouse robot exceed the limits and whether the placement operation is covered by the operating range of the warehouse robot. Pay special attention to checking whether the upper layer goods block the operating space of the lower layer goods. If there is obstruction, adjust the placement order or position of the corresponding goods.

[0023] Example 10: See Figure 1 This embodiment is based on the above embodiment. In step B2, the dynamic process risks include: Step B21: Check if the warehouse robot's movement trajectory is reachable and if there are no joint over-limits. Step B22: Are there any collisions between the goods and the warehouse robot, or between the goods themselves? Step B23: Does the stacking of goods meet stability constraints (no risk of tipping over)?

[0024] Example 11: See Figure 1 This embodiment is based on the above embodiment. In step C2, it is predefined that the wide channel means that it does not directly store goods, but is only used for the movement and path switching of warehouse robots. The number of such channels is small and the width is wide enough to accommodate four warehouse robots. The narrow channel means that it is directly associated with the core area of ​​goods storage and picking. The number of such channels is large and the width is narrow, and the width can only accommodate one warehouse robot.

[0025] Example 12: See Figure 1 This embodiment is based on the above embodiment. In step C3, this embodiment focuses on a dual-block layout e-commerce B2C warehouse. The warehouse is equipped with three 2m wide aisles, namely the front, middle and back, and ten 0.7m wide narrow aisles, namely front blocks 1-5 and back blocks 6-10. It can process an average of 500 orders per day and mixes fragile cosmetics, heavy small appliances and regular daily necessities. It is equipped with four warehouse robots with a load of ≤10kg. The narrow aisles only support one-way passage. Key requirements constraints include: fragile items must be picked first; heavy small appliances should be picked after narrow aisles 7-10; four robots should work together without conflicts in narrow aisles; single order route ≤ 80m; emergency order response time ≤ 15 minutes; and the middle wide aisle should only be used for cross-block switching. Using multi-dimensional weighted cosine similarity technology, the algorithm dataset was retrieved. Each warehouse layout type was labeled with 12 feature dimensions. Weights were assigned according to layout matching (0.3) and constraint matching (0.7). The retrieval results showed that the improved dynamic programming algorithm with priority constraints ranked first with a similarity of 0.92, becoming the final adaptation solution.

[0026] Example 13: See Figure 1 This embodiment is based on the above embodiment. In step C3, this embodiment focuses on an industrial B2B parts warehouse. The warehouse has a single-block layout and is equipped with two wide aisles of 2.2m each, namely the front and rear aisles, and 12 narrow aisles of 0.8m each. The narrow aisles are numbered 1-8 for storing standard fasteners and numbered 9-12 for storing heavy machinery parts. It processes an average of 200 orders from manufacturing customers per day. It mixes precision sensors (fragile and lightweight), heavy bearings (heavy goods), and standard bolts and nuts (small goods). It is equipped with two warehouse robots with a load of ≤15kg. The narrow aisles support two-way passage but require real-time obstacle avoidance. Key requirements and constraints include: precision sensors must be picked first and have separate routes planned; heavy bearings must be picked in the narrow aisles 9-12; two robots must work together without collisions; the route for a single order must be ≤90m; emergency order response time must be ≤20 minutes; and the wide aisle at the back end must be used only for heavy goods transportation (small items must not be occupied by robots). The algorithm dataset was retrieved using multi-dimensional weighted cosine similarity technology. Eight feature dimensions were labeled for each type of warehouse layout. Weights were assigned according to the following weights: fragile item protection adaptation weight 0.3, bidirectional obstacle avoidance weight 0.25, and emergency replenishment response weight 0.15. The retrieval results showed that the improved dynamic programming algorithm with bidirectional obstacle avoidance constraints ranked first with a similarity of 0.65, becoming the final adaptation solution.

[0027] It should be noted that, in this document, 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 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 process, method, article, or apparatus.

[0028] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

[0029] The present invention and its embodiments have been described above. This description is not restrictive, and the accompanying drawings are only one embodiment of the present invention; the actual structure is not limited thereto. In conclusion, if those skilled in the art are inspired by this description and design similar structures and embodiments without departing from the spirit of the invention, such designs should fall within the protection scope of the present invention.

Claims

1. An automated warehouse logistics management system, characterized in that, It includes a warehouse palletizing generation module, a refined scenario verification module, and a back-end logistics adaptive decision-making module; The warehouse palletizing generation module adopts a warehouse palletizing generation method and uses a genetic algorithm to generate a goods placement scheme for palletizing, thereby realizing digital model-driven end-to-end palletizing risk control and conducting preliminary verification of the goods placement scheme. The refined scenario verification module adopts a refined scenario verification method to screen goods placement schemes from a global perspective and perform full-process verification of the final state of palletizing, thereby achieving dual verification of goods placement. The back-end logistics adaptive decision-making module uses a back-end logistics adaptive decision-making method to classify the current warehouse layout and automatically search for picking scheduling strategies that are suitable for the current warehouse picking needs.

2. The automated warehouse logistics management system according to claim 1, characterized in that, The warehouse palletizing generation method specifically includes the following steps: Step A1: Parameter initialization configuration, specifically, input... The system includes high-precision digital models of 3D warehouse robots and cargo, records stacking rules for each type of cargo, sets operational space margins based on the model of the warehouse robot and the size of the end effector, and clarifies optimization goals. Step A2: Initial population generation, randomizing cargo placement schemes; Step A3: Iterative optimization and lightweight verification, used for dynamic filtering during the optimization process. Specifically, prioritizing the basic principles of heavy-to-light and large-to-small, candidate sequences for cargo placement are selected using a genetic algorithm through selection, crossover, and mutation operations. This simplifies the high-precision digital model of the 3D warehouse robot, obtaining a lightweight digital model that retains only the core data of the warehouse robot without graphics rendering. During the iterative process of the genetic algorithm, it verifies in real time whether the coordinates of the cargo in the placement scheme are within the workspace of the warehouse robot, and whether the placement posture causes interference between the warehouse robot body and the cargo. The core data includes joint torque limits, reachable radius, and maximum load mass of the end effector. Each position in the candidate sequence contains backup candidate cargo. Each cargo placement scheme is verified by the lightweight digital model. If the first... If a cargo meets the accessibility and collision-free criteria, its placement and orientation are immediately locked, and the process proceeds to recursively verifying the next cargo. If a cargo does not meet the accessibility and collision-free criteria, the next cargo is adjusted. After verifying the position and orientation of the goods in the order of priority, check again. If the result of the second check is still not satisfactory, switch to the backup candidate goods for re-checking. If the first When all positions and orientations of all alternative candidate goods in the order of priority cannot meet the requirements, i.e., the current order sequence is infeasible, the recursive algorithm automatically backtracks to the next step. Priority goods, readjusted The order of placement and orientation of goods is determined, then the current item is returned to its position. The goods in sequence are re-verified until the next item. A feasible solution is found by sequence, where the solution represents the locked first... The placement and orientation of the goods are ordered, and the solution that maximizes the optimization objective is retained. The optimal solution is then selected for the next round of selection. This process is repeated iteratively until the optimization objective converges. Step A4: Optimal solution output, specifically, outputting the clear order of goods placement, the three-dimensional coordinates and orientation of each goods placement, forming a palletizing solution that can be directly executed.

3. The automated warehouse logistics management system according to claim 2, characterized in that, The refined scenario verification method specifically includes the following steps: Step B1: Set up the simulation environment, including the palletizing scheme, The high-precision digital model of the 3D warehouse robot and the digital model of the goods are imported into the simulation engineering software to restore the real warehouse storage scene, including the location of obstacles. Step B2: Refined scenario verification, used for simulation operation and real-time verification. According to the order, position and posture of goods placement in the palletizing scheme, the complete actions of picking, moving and placing that cannot be covered by the lightweight digital model are gradually simulated by simulating the palletizing process of the warehouse robot. Dynamic process risks are identified and the feasibility of the palletizing scheme is marked. Step B3: Quantitative index output. After the simulation is completed, the simulation engineering software automatically counts and records key indicators, which consist of space utilization and palletizing completion time. Step B4: Iterative correction of the solution. If the palletizing solution is marked as infeasible, return to step A3 of the warehouse palletizing generation module for re-optimization. If it is feasible and the key indicators meet the requirements, the final palletizing solution is output.

4. The automated warehouse logistics management system according to claim 3, characterized in that, The adaptive decision-making method for backend logistics specifically includes the following steps: Step C1: Construct an algorithm dataset for solving the optimal path based on the Traveling Salesman Problem; Step C2: Predefined, specifically, the wide aisle refers to an aisle that does not directly store goods, but is only used for the movement and path switching of warehouse robots, and is large enough to accommodate at least two warehouse robots; the narrow aisle refers to an aisle that is directly associated with the core area of ​​goods storage and picking, and can only accommodate one warehouse robot. Step C3: Define the current warehouse layout, collect the arrangement of wide and narrow aisles in the current warehouse, classify the current warehouse layout to obtain the warehouse layout type, collect the current warehouse logistics requirements, obtain the traveling salesman problem (TSP) requirement constraints, and retrieve the picking and scheduling strategy suitable for the current warehouse from the algorithm dataset based on the warehouse layout type and the TSP requirement constraints.

Images