Intelligent three-dimensional warehouse multi-pile stacker order picking distribution method and system

Abstract

The present application relates to the technical field of automated stereoscopic warehouse logistics scheduling, and proposes an intelligent stereoscopic warehouse multi-stack stacker order allocation method and system, which comprises: multi-stack stacker order allocation problem description, setting model basic assumption and parameter description, constructing a multi-stack stacker order allocation mathematical model, designing an improved NSGA-II optimization algorithm, solving the model according to the actual example of the warehouse and verifying the effect, obtaining the final optimized multi-stack stacker order allocation scheme, and designing a stereoscopic warehouse picking management system, which can effectively solve the resource idling and load problem in the original strategy, significantly improve the overall efficiency of the system, and realize the overall performance improvement of the warehouse operation.

Description

Technical Field

[0001] This invention relates to the field of automated warehouse logistics scheduling technology, and in particular to a method and system for allocating picking orders for intelligent automated warehouses with multiple stacker cranes. Background Technology

[0002] In the context of global economic integration, logistics and transportation account for an increasingly significant proportion of the costs for manufacturing enterprises, exceeding 50%. Faced with fierce market competition, optimizing storage and transportation efficiency and maximizing the use of space resources have become goals pursued by major companies. As a core component of modern logistics systems, the level of intelligence in automated warehouses directly impacts the development of the entire logistics industry. In warehousing operations, efficient picking route planning not only significantly improves logistics efficiency but also effectively reduces logistics costs, thereby enhancing the overall economic benefits of enterprises. This optimization not only helps enterprises gain a competitive edge in the market but also lays a solid foundation for their long-term development.

[0003] Automated storage and retrieval systems (AS / RS), as a highly efficient and intelligent warehousing system, consist primarily of high-rise racking, automated conveyor systems, inbound and outbound workstations, stacker cranes, and advanced operation and control systems. In upstream raw material warehousing applications, these systems handle the storage and turnover of fundamental bulk raw materials. These raw material orders are typically characterized by "large quantities, few categories," meaning downstream manufacturers often need to release large quantities of the same type of material at once to meet continuous production demands. This business characteristic places extremely high demands on the outbound operation efficiency of AS / RS, particularly posing a significant challenge to the scheduling strategies for the coordinated operation of multiple stacker cranes. Summary of the Invention

[0004] Based on the background technology described above, this invention provides a method and system for allocating picking orders for multiple stacker cranes in an intelligent automated warehouse. The invention first provides a basic description of the picking path problem for multiple stacker cranes, sets the basic assumptions and parameters of the model, constructs a mathematical model for allocating picking orders for multiple stacker cranes, designs an improved NSGA-II optimization algorithm, and, based on a warehouse example, uses the improved NSGA-II optimization algorithm to solve the model and performs effect verification analysis to obtain the final optimized picking order allocation scheme for multiple stacker cranes.

[0005] The technical means applied in this invention include:

[0006] A method and system for allocating picking orders using multiple stacker cranes in an intelligent automated warehouse includes the following steps:

[0007] S1. Description of the picking path problem for multiple stacker cranes: Define the core problem in the scenario of collaborative operation of multiple stacker cranes. Given a batch of goods to be shipped and their location coordinates, how to reasonably allocate these picking tasks to each stacker crane to minimize the total system completion time and avoid warehouse relocation operations as much as possible.

[0008] S2. Construction of a Mathematical Model for Picking Order Allocation with Multiple Stacker Cranes: To study the efficient operation of automated storage and retrieval systems, basic assumptions are set regarding shelves, stacker cranes, and operational processes. In addition, the symbols and parameters used in the model are defined and explained, and the time for transferring goods and picking single items is calculated. With the goal of minimizing the total time required to complete the entire order, unique allocation constraints for goods and domain constraints for decision variables are set to construct a mathematical model for picking order allocation with multiple stacker cranes.

[0009] S3. Improved NSGA-II Optimization Algorithm Design: To solve the model, a cloud drift optimization mechanism is introduced to form an enhanced architecture with NSGA-II. This improved algorithm combines the multi-objective optimization framework of NSGA-II with the intelligent search mechanism of cloud drift, which significantly improves the quality and convergence speed of the solution while maintaining population diversity. It is suitable for the stacker crane path optimization problem with complex constraints proposed in this paper.

[0010] S4. Model Solving and Effect Verification: Set up shelf and goods information, use Python to perform simulation calculations, and obtain the inbound and outbound order table; use the improved NSGA-II algorithm to solve the problem based on the outbound order of the case, and obtain the multi-stacker crane picking order allocation scheme; compare and verify the output scheme with the original scheme to determine whether the optimized task allocation scheme has achieved the expected optimization goal.

[0011] Design of automated warehouse picking management system: In order to ensure the load balance of picking by multiple stacker cranes and the visualized management of warehouse status, we will develop a software that is easy to operate, stable in operation, and can quickly optimize calculations, and carry out overall architecture design and functional design.

[0012] A method and system for allocating picking orders for multiple stacker cranes in an intelligent automated warehouse is characterized by: using an improved NSGA-II algorithm to achieve effective allocation of picking orders for multiple stacker cranes during automated warehouse operations, solving the problems of equipment load and idleness in warehousing resources, and improving the overall efficiency of warehousing operations.

[0013] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0014] 1. This invention constructs a mathematical model for order allocation of multiple stacker cranes to minimize the total time required to complete the entire order, and fully considers the constraints of goods allocation and decision variable domains.

[0015] 2. The improved cloud drift NSGA-II algorithm designed in this invention can accurately calculate the shortest path from the starting position of the stacker crane to the target goods, further shortening the picking time, improving the efficiency of warehousing operations, and providing decision-making reference for the intelligent development of automated three-dimensional warehousing for enterprises. Instruction manual illustrations

[0016] Figure 1 This is a flowchart of the present invention.

[0017] Figure 2 This is the algorithm flowchart.

[0018] Figure 3 This is a graph showing the trend of the objective function.

[0019] Figure 4 Load diagrams of stacker cranes before and after optimization.

[0020] Figure 5 This is a screenshot of the system interface.

[0021] Figure 6 For the picking task list. Detailed Implementation

[0022] The technical solution and the flowchart of the present invention will be explained in detail below.

[0023] like Figure 1 As shown, a method and system for allocating picking orders for intelligent automated warehouses with multiple stacker cranes are proposed, including:

[0024] S1. Description of the picking path problem for multiple stacker cranes: Define the core problem in the scenario of collaborative operation of multiple stacker cranes. Given a batch of goods to be shipped and their location coordinates, how to reasonably allocate these picking tasks to each stacker crane to minimize the total system completion time and avoid warehouse relocation operations as much as possible.

[0025] In specific implementation, taking the multi-stacker picking of Company X as an example, the multi-stacker picking order allocation analysis addresses the uneven loading of multiple stacker cranes caused by unreasonable task allocation. It also considers how to anticipate and avoid unnecessary material handling operations during task allocation, aiming to reasonably allocate goods to stacker cranes and achieve efficient picking under the conditions of the quantity of goods and the operating characteristics of stacker cranes.

[0026] S2. Mathematical model for order allocation of multiple stacker cranes is constructed to set basic assumptions about shelves, stacker cranes, and work processes in order to study the efficient operation of automated storage and retrieval systems.

[0027] In specific implementation, the basic assumptions include:

[0028] The warehouse is equipped with a row of standard-sized shelves, with each storage location having the same dimensions. The warehouse's inbound and outbound entrances are located at opposite ends of the shelves, and each aisle is equipped with a stacker crane responsible for loading and unloading operations. These stacker cranes are double-reach stacker cranes, capable of handling picking operations from two rows of shelves on each side. The stacker cranes can move independently in both horizontal and vertical directions at a fixed speed. Ignoring the start-up and braking times, the loading and unloading time at any storage location is constant and does not change due to changes in the picking order. In each operation, the stacker crane can complete all inbound and outbound requirements for a specific type of goods in an order in one go. It is assumed that each storage location on the shelves is of the same size, and the quantity of goods at each location meets safety stock requirements. When an order is placed, the warehouse management system immediately assigns and notifies the specific storage location coordinates for inbound and outbound operations. All stacker cranes used are of the same model, and operations are performed facing the front of the shelves. Inbound and outbound operations are conducted separately, assuming the stacker crane's initial position is the outbound entrance.

[0029] Define and explain the symbols and parameters used in the model.

[0030] In practical implementation, the symbols and parameters used in the model are defined and explained as shown in Table 1: Table 1 Parameter Settings parameter illustrate K Number of stacker cranes, Kr∈{1, 2, 3, 4, ...} M Picking order goods collection <![CDATA[v y ]]> Stacker crane horizontal movement speed <![CDATA[v z ]]> Stacker crane vertical movement speed parameter illustrate L Storage compartment length and width <![CDATA[(r m ,x m ,y m ,z m )]]> <![CDATA[Goods m lane, row, column, and layer coordinates]]> <![CDATA[t m ]]> Single operation time for stacker crane picking goods m <![CDATA[t y ]]> Transfer time <![CDATA[(y e ,With e )]]> Column and layer coordinates of the nearest empty storage location <![CDATA[t p ]]> Single picking operation time <![CDATA[R f ]]> <![CDATA[Internal roadway, R f ∈ {2, 3, 6, 7, 10, 11.....}]]> Rs External passageways, Rs∈{1, 4, 5, 8, 9, 12, ...}

[0031] Calculate the time for warehouse transfer operations and the time for picking individual items.

[0032] In practice, the total transfer time in the warehouse transfer operation is expressed as:

[0033]

[0034] When goods in an external aisle are blocked by goods in an internal aisle, the blocking goods must first be moved to the nearest empty storage location before the picking operation can continue. This value is 0 when the storage location is an internal location or is not blocked by other goods.

[0035] In practice, the time required for a stacker crane to complete one picking operation of goods is expressed as:

[0036]

[0037] The operation time varies depending on whether the cargo is located in an internal or external aisle.

[0038] Construct a mathematical model for order allocation for multiple stacker cranes.

[0039] In practical implementation, the objective is to minimize the total time required to complete the entire order, and the objective function can be expressed as:

[0040]

[0041] Set the constraints for the mathematical model of the multi-stacker picking order allocation.

[0042] In specific implementation, the constraints include:

[0043] Each item in the order must be assigned, and only one stacker crane can be responsible for picking it:

[0044]

[0045] Decision variable X mk The value range constraint means that the relationship between a piece of goods m and a stacker crane k can only be "yes" or "no", that is, the goods are either assigned to this stacker crane (value 1) or not assigned (value 0):

[0046]

[0047] S3. Improved NSGA-II optimization algorithm design: The algorithm is improved by combining the multi-objective optimization framework of NSGA-II with the intelligent search mechanism of cloud drift.

[0048] NSGA-II achieves global exploration through population evolution via genetic operations, ensuring the diversity of solutions. Cloud drift optimization, on the other hand, employs an elite-guided directional search strategy, intelligently perturbing around elite solutions to effectively enhance population diversity and avoid premature convergence. The improved NSGA-II optimization algorithm significantly improves the quality and convergence speed of solutions while maintaining population diversity, making it suitable for the stacker crane path optimization problem with complex constraints proposed in this invention.

[0049] In practical implementation, the NSGA-II algorithm was improved, and a composite coding scheme based on cargo-location-stacking crane allocation was adopted in the coding method:

[0050]

[0051] Where g j Let (aj, rj, cj, lj) be the cargo number, (aj, rj, cj, lj) be the cargo location coordinates, and s be the cargo number. j This is the stacker crane number.

[0052] In specific implementation, the solution steps of the improved cloud drift NSGA-II algorithm are as follows: Figure 2 As shown.

[0053] Perform algorithm effectiveness analysis.

[0054] In practice, the same outbound order under the same environment was selected, and three representative optimization algorithms were chosen as controls for a multi-algorithm comparison experiment to obtain a performance comparison of each algorithm:

[0055] Table 2 Algorithm Performance Comparison Results Model optimal value Worst value Standard deviation Average value of objective function Optimization rate NSGA-II 221 232 5.643 228.05 1.92% MOPSO 235 251 9.523 245.91 8.97% ACS 220 246 6.996 232.45 3.69% CDO-NSGA-II 214.0 229.0 5.303 223.85 /

[0056] In practical implementation, the scalability of the algorithm and its performance under different problem sizes were further verified. Three types of cases—large, medium, and small—were set up, and each type of case was run independently 20 times in the same computing environment to eliminate the bias caused by single-run randomness. The statistical results are shown in Table 3.

[0057] Table 3. Results of Algorithm Scalability Test Example size optimal value Worst value Standard deviation Average value of objective function small-scale 94 102 4.432 98.05 medium-sized 214.0 229.0 5.303 223.85 large-scale 528 544 5.064 536.55

[0058] S4. Model Solving: Set up shelf and goods information, use Python to perform simulation calculations based on order data, and obtain the inbound and outbound order table; use the improved NSGA-II algorithm to solve the problem based on the case outbound order to obtain the multi-stacker crane picking order allocation scheme.

[0059] In the specific implementation, a certain order from Company X on a certain day was selected, with a total of 27 items to be shipped. The verification study was conducted based on the company's actual automated warehouse, and simulation calculations were performed using Python.

[0060] The inbound and outbound order table is generated.

[0061] Table 4. List of Goods to be Shipped Goods Name quantity 61000842 3 61000828 9 61000821 6 61000814 5 61000801 1 61000786 3

[0062] The improved NSGA-II optimization algorithm was used to solve and optimize the model.

[0063] By using the improved algorithm for optimization, the termination condition was reached after 60 iterations, and the stacker crane's running time was 214 seconds.

[0064] The trend of the target value change is obtained as follows Figure 3 As shown.

[0065] The detailed picking schemes before and after the optimization strategy are shown in the table below.

[0066] Table 5 Picking Plan

[0067] Verify the effectiveness based on the solution results.

[0068] In practice, the original picking order allocation scheme and the optimized scheme for multi-stacker crane warehousing operations at Company X will be compared. The stacker crane load diagrams before and after optimization are shown below. Figure 4 As shown.

[0069] As shown in the figure, compared with the original scheme, the optimized picking strategy has achieved significant improvement in load balancing.

[0070] The working time and equipment utilization rate of each stacker crane are shown in the table below.

[0071] Table 6 Comparison of Picking Strategy Optimization

[0072] As shown in the table, before optimization, there were significant differences in the working time of each stacker crane, with equipment utilization rates ranging from 80% to 100%, indicating a clear load imbalance. After strategy optimization, the working time of the four stacker cranes became more balanced, the maximum working time difference was reduced from 36 seconds to only 5 seconds, and the utilization rate of all equipment increased to over 95%.

[0073] The picking time was analyzed based on the experimental results.

[0074] Table 7 Comparison of Objective Function Values function value before optimization Optimized function value Difference before and after optimization Optimization ratio Picking time 247 214 33 13.36%

[0075] The data in the table shows that the optimization strategy proposed in this invention has achieved certain results in improving operational efficiency. The maximum picking time of the system was reduced from 247 seconds before optimization to 227 seconds after optimization, representing a relative optimization of 13.36%.

[0076] This improvement fully demonstrates the effectiveness of the optimization strategy in balancing the workload of stacker cranes and reducing operational bottlenecks, laying a solid foundation for improving the overall efficiency of warehouse operations.

[0077] Design of automated warehouse picking management system: In order to ensure the load balance of picking by multiple stacker cranes and the visualized management of warehouse status, we will develop a software that is easy to operate, stable in operation, and can quickly optimize calculations, and carry out overall architecture design and functional design.

[0078] In practice, the system's main interface adopts a tabbed layout, consisting of a menu bar, tab labels, and a function panel from top to bottom. Each tab contains controls arranged according to their function. The system includes tabs for correlation mining, location allocation, outbound picking, and warehouse status. The system interface is displayed as follows: Figure 5 .

[0079] In practical implementation, the functional design covers basic functions and core business functions. Basic functions include login, data import, result saving, interface refresh and feedback, and exit. Core business functions include association rule mining, location and picking optimization, and warehouse status visualization, realizing the integration of functions throughout the entire intelligent three-dimensional warehousing process.

[0080] The system runs stably and is easy to operate. Some of the results are shown below. Figure 6 .

[0081] The system design verified the operability and systematic implementation capability of the multi-stacking crane picking order allocation optimization strategy, and ensured the smooth implementation and continuous optimization of the optimization plan through information technology.

[0082] The above detailed description of the embodiments of the present invention is an illustrative example of the methods and steps of the present invention. Those skilled in the art can supplement and modify the specific embodiments within the scope of the definition of the present invention, and the scope of the present invention is defined by the appended claims.

Claims

1. A method for allocating picking orders using a multi-stacking crane in an intelligent automated warehouse, comprising the following steps: S1. Description of the multi-stacking crane picking order allocation problem: Define the core problem in the scenario of multi-stacking crane collaborative operation. Given a batch of goods to be shipped and their location coordinates, how to reasonably allocate these picking tasks to each stacker crane to minimize the total system completion time and avoid warehouse relocation operations as much as possible. S2. Construction of a mathematical model for order allocation by multiple stacker cranes: To study the efficient operation of automated storage and retrieval systems, basic assumptions are set for racks, stacker cranes, and work processes. In addition, the symbols and parameters used in the model are defined and explained, and the time for warehouse transfer operation and single item picking operation is calculated. With the goal of minimizing the total time required to complete the entire order, a mathematical model for order allocation for multiple stacker cranes is constructed by setting unique cargo allocation constraints and decision variable domain constraints. S3. Improved NSGA-II Optimization Algorithm Design: To solve the model, a cloud drift optimization mechanism is introduced to form an enhanced architecture with NSGA-II. This improved algorithm combines the multi-objective optimization framework of NSGA-II with the intelligent search mechanism of cloud drift, which significantly improves the quality and convergence speed of the solution while maintaining population diversity. It is suitable for the stacker crane path optimization problem with complex constraints proposed in this paper. S4. Model Solving and Effect Verification: Set up shelf and goods information, use Python to perform simulation calculations, and obtain the inbound and outbound order table; Using the improved NSGA-II algorithm, a multi-stacker crane picking order allocation scheme is obtained based on the outbound order case. The output solution is compared and verified with the original solution to determine whether the optimized task allocation solution has achieved the expected optimization goal.

2. The intelligent automated warehouse multi-stacking crane picking order allocation method described in claim 1, characterized in that: By employing an improved NSGA-II algorithm, this paper achieves effective allocation of picking orders from multiple stacker cranes during automated warehouse operations, resolving issues of equipment load and idleness in warehousing resources and improving overall warehousing efficiency. The multi-stacker crane picking order allocation analysis addresses the uneven load distribution among stacker cranes caused by unreasonable task allocation. It also considers how to anticipate and avoid unnecessary material handling operations during task allocation, aiming to rationally allocate goods to stacker cranes and achieve efficient picking under conditions such as the quantity of goods and the operational characteristics of the stacker cranes.

3. The intelligent automated warehouse multi-stacking crane picking order allocation method described in claim 1, characterized in that, The model assumptions include: The warehouse is equipped with a row of standard-sized shelves, with each storage location having the same dimensions. The warehouse's inbound and outbound entrances are located at opposite ends of the shelves, and each aisle is equipped with a stacker crane responsible for loading and unloading operations. These stacker cranes are double-reach stacker cranes, capable of handling picking operations from two rows of shelves on each side. The stacker cranes can move independently in both horizontal and vertical directions at a fixed speed. Ignoring the start-up and braking times, the loading and unloading time at any storage location is constant and does not change due to changes in the picking order. In each operation, the stacker crane can complete all inbound and outbound requirements for a specific type of goods in an order in one go. It is assumed that each storage location on the shelves is of the same size, and the quantity of goods at each location meets safety stock requirements. When an order is placed, the warehouse management system immediately assigns and notifies the specific storage location coordinates for inbound and outbound operations. All stacker cranes used are of the same model, and operations are performed facing the front of the shelves. Inbound and outbound operations are conducted separately, assuming the stacker crane's initial position is the outbound entrance.

4. The intelligent automated warehouse multi-stacking crane picking order allocation method described in claim 1, characterized in that, The symbols and parameter definitions and descriptions used in the model include: Table 1 Parameter Settings 5. The intelligent automated warehouse multi-stapler picking order allocation method and system described in claim 1, characterized in that, The total transfer time in a warehouse transfer operation is expressed as follows: When goods in an external aisle are blocked by goods in an internal aisle, the blocking goods must first be moved to the nearest empty storage location before the picking operation can continue. This value is 0 when the storage location is an internal location or is not blocked by other goods. The time required for a stacker crane to complete one picking operation of goods is expressed as: The operation time varies depending on whether the cargo is located in an internal or external aisle.

6. The intelligent automated warehouse multi-stacking crane picking order allocation method described in claim 1, characterized in that, The objective function, with the goal of minimizing the total time required to complete the entire order, can be expressed as: The function first calculates the independent picking time for each item, which is the round-trip travel time of the stacker crane from its inbound / outbound station to the corresponding storage location. Then, it sums up the picking times for all items assigned to the same stacker crane to obtain the total load for each crane. Ultimately, the goal is to find a picking scheme from all possible item allocation options that minimizes the maximum load (i.e., the latest completion time) on each stacker crane.

7. The intelligent automated warehouse multi-stacking crane picking order allocation method according to claim 1, characterized in that, The constraints include: Each item in the order must be assigned, and only one stacker crane can be responsible for picking it: Decision variable X mk The value range constraint means that the relationship between a piece of goods m and a stacker crane k can only be "yes" or "no", that is, the goods are either assigned to this stacker crane (value 1) or not assigned (value 0):

8. The intelligent automated warehouse multi-stacking crane picking order allocation method according to claim 1, characterized in that, The improved NSGA-II algorithm design specifically includes: By combining the multi-objective optimization framework of NSGA-II with the intelligent search mechanism of cloud drift, an improved NSGA-II algorithm is designed, and a composite coding scheme based on cargo-location-stacking crane allocation is adopted in the coding method: in, Assign a cargo number, For cargo coordinates, This is the stacker crane number.

9. A method for allocating picking orders for an intelligent automated warehouse with multiple stacker cranes according to claim 1, characterized in that, The model solving and effect verification specifically include: A practical example analysis was conducted to solve the model and obtain a multi-stacking crane picking order allocation scheme. Based on the results, the optimization strategy was verified and analyzed, proving its effectiveness in balancing the workload of stacker cranes and reducing operational bottlenecks, thus laying a solid foundation for improving the overall efficiency of warehouse operations.

10. A smart automated warehouse multi-stacking crane picking order allocation system, characterized in that, To ensure load balancing and visualized management of warehouse status during multi-stall crane picking, a user-friendly, stable, and fast optimization software was developed, along with overall architecture and functional design. A "Self-Service Warehouse Picking Management System" was designed and developed, comprising four modules: correlation mining, location allocation, outbound picking, and warehouse status. The feasibility and systematic implementation capabilities of the optimization strategies were verified.

Images