Intelligent warehouse logistics sorting system and method

Abstract

The invention relates to the technical field of intelligent warehouse logistics, and discloses an intelligent warehouse logistics sorting system and method, and the system comprises a multi-dimensional data collection module, a global state perception and simulation model construction module, a demand prediction analysis module, a task optimization and resource configuration module, and an execution control and feedback adjustment module. The system realizes dynamic demand prediction, task optimization division and resource adaptation through a distributed sensing network, a three-dimensional simulation model and a deep learning algorithm, and ensures material storage safety in combination with environment adaptive regulation and control. The warehouse sorting efficiency can be improved, the resource allocation is optimized, the environmental adaptability is enhanced, and the method is suitable for modern intelligent warehouse scenes.

Description

Technical Field

[0001] This invention belongs to the field of intelligent logistics and warehouse automation technology, specifically an intelligent warehousing and logistics sorting system and method. Background Technology

[0002] With the rapid development of e-commerce and intelligent manufacturing, the complexity and scale of warehousing and logistics systems are constantly expanding. Traditional warehousing management models face numerous challenges in meeting the demands for efficient and accurate material sorting and distribution. Modern warehousing systems need to process massive amounts of multi-source heterogeneous data, including order information, inventory status, and environmental parameters, and respond to dynamically changing business needs in real time. However, current warehousing systems still exhibit inefficiencies and limited levels of intelligence in areas such as task scheduling, resource allocation, and environmental control.

[0003] Existing warehousing systems typically rely on static rules or single algorithms for task allocation and path planning, lacking real-time awareness of the overall status and making it difficult to effectively respond to emergencies or predict future demand. Furthermore, in managing the storage environment of sensitive materials, most existing systems only provide passive monitoring functions, unable to proactively adjust storage conditions to ensure material quality. Therefore, there is an urgent need for a warehousing and logistics sorting system that can integrate multi-source data, optimize path planning, dynamically adjust storage layout, and achieve intelligent prediction, in order to improve overall operational efficiency and adaptability. Summary of the Invention

[0004] The purpose of this invention is to provide a centralized supply chain management system for operating rooms based on robot-assisted automation, in order to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides an intelligent warehousing and logistics sorting system, comprising: The multi-dimensional data acquisition module is configured to acquire and integrate dynamic and static information related to warehousing operations from multiple heterogeneous sources. The global state perception and simulation model construction module is configured to receive data provided by the multi-dimensional data acquisition module and build and maintain a three-dimensional simulation model that is synchronized with the physical warehouse in real time. The demand forecasting and analysis module is configured to make probabilistic estimates of material outbound demand based on historical business data and future planning information, and generate a future demand distribution map. The task optimization and resource allocation module is configured to receive new outbound orders and, in conjunction with the real-time status information of the 3D simulation model and the future demand distribution map, dynamically divide the outbound orders into several task groups, and allocate appropriate execution resources and optimal paths to each task group. The execution control and feedback adjustment module is configured to convert the instructions generated by the task optimization and resource configuration module into specific operation commands for the execution device, and to monitor and dynamically adjust the execution process in real time.

[0006] Preferably, the multi-dimensional data acquisition module specifically includes: The warehouse management interface unit is configured to retrieve new outbound order information at fixed time intervals; The enterprise resource scheduling interface unit is configured to extract historical sales records, production plans, and marketing activity information to support demand forecasting. The distributed sensor network is configured to be deployed in a grid-like manner within the physical space of the warehouse to collect environmental parameters, real-time location and operational status information of physical entities.

[0007] Preferably, the distributed sensor network further includes: Temperature and humidity sensor arrays are installed on pre-set grid nodes in the warehouse, with the temperature sensor having a measurement accuracy of no less than ±0.3℃ and the humidity sensor having a measurement accuracy of no less than ±3%RH. The ultra-wideband positioning system consists of positioning base stations and positioning tags. It is configured to provide the three-dimensional spatial coordinates of the positioning tags attached to automated transport equipment, containers or operators at an update frequency of not less than 15 Hz, with a positioning error of not more than 20 centimeters. An ultrasonic ranging sensor, embedded inside a transport container, is designed to calculate the container's space utilization rate in real time by measuring the height distribution of materials inside the container.

[0008] Preferably, the three-dimensional simulation model constructed and maintained by the global state perception and simulation model construction module includes: A dynamic weighted warehouse topology model, where nodes represent functional areas within the warehouse, edges represent connecting paths between functional areas, and the weights of the edges are adjusted in real time based on the dynamic flow coefficients calculated from the real-time density and average speed of automated transport equipment on the path. The full resource status matrix is ​​set to record the status of each automated transportation device and each transportation container in real time; The material lifecycle state vector is set to record the precise storage location, quantity, batch information, and cumulative environmental impact value of each material in the inventory, which is obtained by continuous integration calculation based on the real-time environmental data of its location and exposure duration.

[0009] Preferably, the dynamic weighted warehouse topology model is stored in a relational database; the nodes represent functional areas such as storage locations, charging stations, operating platforms, or inbound / outbound channels; the basic weight of the edges is physical distance; and the update cycle of the dynamic flow coefficient is no more than 3 seconds.

[0010] Preferably, the three-dimensional simulation model also includes a three-dimensional environmental distribution map generated by interpolation algorithm from the data collected by the temperature and humidity sensor array. This map represents the real-time temperature and humidity conditions at any coordinate point in the warehouse in the form of a voxel grid.

[0011] Preferably, the demand forecasting and analysis module adopts a deep learning model based on convolutional neural networks. The input layer of the model receives historical order demand time series data in terms of material categories and corresponding external variables such as market activities and seasonal factors. The model takes the actual demand data of the previous period and the production plan in the future preset period as input and outputs a probabilistic demand forecast vector. This vector gives the probability and expected quantity of each material being demanded in each time window in the future preset period, and generates a demand distribution map indicating the future picking pressure in different areas of the warehouse.

[0012] Preferably, the task optimization and resource allocation module dynamically partitions orders by running a multi-objective particle swarm optimization algorithm; the multi-objective particle swarm optimization algorithm evaluates each order partitioning scheme using a preset multi-objective fitness function; the fitness function is a weighted linear combination function containing at least three of the following terms: A path cost function designed to minimize the total length of planned paths for all task groups; A container efficiency function designed to maximize the space utilization of all containers used. The aim is to minimize the time delay function that exceeds the expected completion time of an order from its committed service level agreement; An impact function designed to minimize the incremental environmental impact of sensitive materials during picking and transportation.

[0013] Preferably, the task optimization and resource allocation module further includes a resource allocation and path planning unit, which is configured to select the final execution scheme from a set of optimal solutions found by the multi-objective particle swarm optimization algorithm, and select suitable automated transportation equipment and transportation containers for each task group in the scheme from the full resource state matrix according to the total weight and total volume of materials in the group, and call the improved A* path search algorithm to calculate a real-time optimal path for the selected automated transportation equipment on the dynamic weight warehouse topology model.

[0014] This invention provides an intelligent warehousing and logistics sorting method, comprising the following steps: The multi-dimensional data acquisition module acquires and integrates dynamic and static information related to warehousing operations from multiple heterogeneous sources; The global state perception and simulation model building module receives data provided by the multi-dimensional data acquisition module and builds and maintains a three-dimensional simulation model that is synchronized with the physical warehouse in real time. The demand forecasting and analysis module uses historical business data and future planning information to probabilistically estimate the outbound demand for materials and generate a future demand distribution map. The task optimization and resource allocation module receives new outbound orders and, in conjunction with the real-time status information of the 3D simulation model and the future demand distribution map, dynamically divides the outbound orders into several task groups and allocates suitable execution resources and optimal paths to each task group. The execution control and feedback adjustment module transforms the instructions generated by the task optimization and resource configuration module into specific operation commands for the execution device, and performs real-time monitoring and dynamic adjustment of the execution process.

[0015] Compared with the prior art, the beneficial effects of the present invention are: 1. Improve warehouse sorting efficiency: By integrating heterogeneous data (orders, environment, location, etc.) through a multi-dimensional data acquisition module and combining it with a real-time 3D simulation model built by a global state perception module, the warehouse's overall state can be accurately controlled. The task optimization module uses a multi-objective particle swarm optimization algorithm to dynamically divide orders and improves the A* algorithm to plan the optimal path, reducing path congestion and invalid operations, significantly improving the overall efficiency of the sorting process and shortening order completion time.

[0016] 2. Optimize resource allocation: The task optimization and resource allocation module matches suitable automated transportation equipment and containers for task groups based on material weight, volume and full resource status matrix. At the same time, it maximizes space utilization through container efficiency function, reduces equipment idleness and container waste, and realizes efficient matching and scheduling of equipment, containers and other resources.

[0017] 3. Enhance environmental adaptability and dynamic response capabilities. The dynamic weighted warehouse topology model of the global state perception module updates the path flow coefficient every 3 seconds. Combined with the real-time monitoring and dynamic adjustment mechanism of the execution control module, it can quickly respond to emergencies such as path congestion and flexibly adjust the operation plan. The real-time environmental distribution map of the 3D simulation model supports accurate perception of the warehouse environment and improves the system's adaptability to dynamic environments. Attached Figure Description

[0018] Figure 1 This is a schematic diagram of the overall architecture of the intelligent warehousing and logistics sorting system of the present invention.

[0019] Figure 2 This is a schematic diagram of the deployment of a distributed sensor network.

[0020] Figure 3A schematic diagram of the 3D simulation model generated by the global state perception and simulation model construction module.

[0021] Figure 4 A flowchart illustrating the workflow of the demand forecasting and analysis module. Figure 5 This is a schematic diagram of task division and path planning for the task optimization and resource allocation module. Detailed Implementation

[0022] This invention provides a specific implementation of an intelligent warehousing and logistics sorting system, in conjunction with the appendix. Figure 1 To be continued Figure 5 A detailed explanation is provided. The system includes a multi-dimensional data acquisition module, a global state perception and simulation model construction module, a demand prediction and analysis module, a task optimization and resource allocation module, an execution control and feedback adjustment module, and an environmental adaptive control module. Each module achieves intelligent warehousing and logistics sorting functions through data interaction and logical connections.

[0023] As attached Figure 1 As shown, the multi-dimensional data acquisition module is the fundamental information source for the entire system. It internally includes a warehouse management interface unit, an enterprise resource scheduling interface unit, and a distributed sensor network. The warehouse management interface unit retrieves new outbound order information from the warehouse management system at fixed time intervals and transmits this information to the global state perception and simulation model construction module and the task optimization and resource allocation module. The enterprise resource scheduling interface unit extracts historical sales records, production plans, and marketing activity information to provide necessary input data for the demand forecasting and analysis module.

[0024] As a preferred embodiment, as shown in the appendix Figure 2 As shown, the distributed sensor network is deployed in a grid-like layout within the warehouse. It comprises a temperature and humidity sensor array, an ultra-wideband positioning system, and ultrasonic ranging sensors. The temperature and humidity sensor array is installed on pre-defined grid nodes in the warehouse, with measurement accuracies of ±0.3℃ and ±3%RH, respectively, for real-time monitoring of warehouse environmental parameters.

[0025] The ultra-wideband positioning system consists of positioning base stations and positioning tags. The tags are attached to automated transport equipment, containers, or operators, with an update frequency of at least 15 Hz and a positioning error of no more than 20 centimeters, used for real-time tracking of the physical location information of entities. Ultrasonic ranging sensors are embedded inside the transport container, calculating space utilization by measuring the height distribution of materials within the container, and transmitting the data to a global state perception and simulation model construction module.

[0026] As a preferred embodiment, as shown in the appendix Figure 3As shown, the global state perception and simulation model building module receives data from the multi-dimensional data acquisition module and builds a three-dimensional simulation model. This model includes a dynamically weighted warehouse topology model, a full resource state matrix, and a material lifecycle state vector.

[0027] The dynamic weighted warehouse topology model is stored in a relational database. Nodes represent functional areas within the warehouse, such as storage locations, charging stations, operating tables, or inbound / outbound channels. Edges represent the connection paths between functional areas. The basic weight is the physical distance, and the dynamic flow coefficient is updated every 3 seconds based on the real-time density and average speed of automated transportation equipment on the path.

[0028] The full resource status matrix records the status of each automated transportation device and each transportation container in real time, including information such as device location, operating status, and container space utilization. The material lifecycle status vector records the precise storage location, quantity, batch information, and cumulative environmental impact value of each material in the inventory, calculated by continuous integration based on real-time environmental data and exposure duration.

[0029] In addition, the three-dimensional simulation model also includes a three-dimensional environmental distribution map generated from data collected by an array of temperature and humidity sensors. This map represents the real-time temperature and humidity conditions at any coordinate point in the warehouse in the form of a voxel grid.

[0030] As a preferred embodiment, as shown in the appendix Figure 4 As shown, the demand forecasting and analysis module uses a deep learning model based on convolutional neural networks to probabilistically estimate material outbound demand. The model's input layer receives historical order demand time series data categorized by material type, along with corresponding external variables such as market activities and seasonal factors. The actual demand data from the previous period and the production plan for the future preset time period are used as inputs. The output is a probabilistic demand forecast vector, providing the probability and expected quantity of each material to be demanded in each time window within the future preset time period.

[0031] The demand forecasting and analysis module generates a future demand distribution map based on probabilistic demand forecasting vectors, indicating the future picking pressure in different areas of the warehouse, and then transmits this distribution map to the task optimization and resource allocation module.

[0032] As a preferred embodiment, as shown in the appendix Figure 5 As shown, the task optimization and resource allocation module realizes the dynamic partitioning of orders by running a multi-objective particle swarm optimization algorithm. The multi-objective particle swarm optimization algorithm uses a preset multi-objective fitness function to evaluate each order partitioning scheme. The fitness function includes path cost function, container efficiency function, time delay function and impact function.

[0033] The path cost function aims to minimize the total length of planned paths for all task groups; the container efficiency function aims to maximize the space utilization of all containers used; the time delay function aims to minimize the time delay between the expected completion time of an order and its committed service level agreement; and the impact function aims to minimize the incremental environmental cumulative impact value of sensitive materials during picking and transportation.

[0034] The task optimization and resource allocation module further includes a resource allocation and path planning unit, which selects the final execution scheme from a set of optimal solutions found by the multi-objective particle swarm optimization algorithm, and selects suitable automated transportation equipment and transportation containers for each task group in the scheme from the full resource state matrix based on the total weight and total volume of the materials in the group.

[0035] The improved A* path search algorithm calculates a real-time optimal path for selected automated transportation equipment on a dynamically weighted warehouse topology model and transmits the path information to the execution control and feedback adjustment module.

[0036] The execution control and feedback adjustment module translates the instructions generated by the task optimization and resource allocation module into specific operation commands for the execution equipment, and monitors and dynamically adjusts the execution process in real time. Through interaction with a distributed sensor network, the module obtains the location information and operating status of the automated transportation equipment in real time, and dynamically adjusts task allocation and route planning based on the actual execution situation. For example, when the dynamic flow coefficient of a certain route increases due to sudden congestion, the execution control and feedback adjustment module will re-invoke the improved A* path search algorithm to calculate a new optimal route and update the operating instructions of the automated transportation equipment.

[0037] The environmental adaptive control module continuously compares the environmental data in the 3D simulation model with the suitable storage condition ranges for each sensitive material defined in the material master data. When an environmental parameter deviates from the safe range, the environmental adaptive control module triggers preset control measures or emergency response mechanisms. For example, when the temperature of a storage location exceeds the suitable range for a sensitive material, the environmental adaptive control module will activate local temperature control equipment or adjust the material storage location to ensure material safety.

[0038] The aforementioned modules achieve the overall operation of the intelligent warehousing and logistics sorting system through data interaction and logical connections. The multi-dimensional data acquisition module provides basic data support; the global state perception and simulation model construction module builds a real-time synchronized 3D simulation model; the demand forecasting and analysis module generates a future demand distribution map; the task optimization and resource allocation module realizes dynamic order allocation and optimal resource configuration; the execution control and feedback adjustment module is responsible for instruction conversion and real-time adjustment; and the environmental adaptive control module ensures the storage security of sensitive materials.

[0039] 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 term "inclusion" or any other variation thereof is 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 a process, method, article, or apparatus.

[0040] 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.

Claims

1. An intelligent warehousing and logistics sorting system, characterized in that, include: The multi-dimensional data acquisition module is configured to acquire and integrate dynamic and static information related to warehousing operations from multiple heterogeneous sources. The global state perception and simulation model construction module is configured to receive data provided by the multi-dimensional data acquisition module and build and maintain a three-dimensional simulation model that is synchronized with the physical warehouse in real time. The demand forecasting and analysis module is configured to make probabilistic estimates of material outbound demand based on historical business data and future planning information, and generate a future demand distribution map. The task optimization and resource allocation module is configured to receive new outbound orders and, in conjunction with the real-time status information of the 3D simulation model and the future demand distribution map, dynamically divide the outbound orders into several task groups, and allocate appropriate execution resources and optimal paths to each task group. The execution control and feedback adjustment module is configured to convert the instructions generated by the task optimization and resource configuration module into specific operation commands for the execution device, and to monitor and dynamically adjust the execution process in real time.

2. The intelligent warehousing and logistics sorting system according to claim 1, characterized in that, The multi-dimensional data acquisition module specifically includes: The warehouse management interface unit is configured to retrieve new outbound order information at fixed time intervals; The enterprise resource scheduling interface unit is configured to extract historical sales records, production plans, and marketing activity information to support demand forecasting. The distributed sensor network is configured to be deployed in a grid-like manner within the physical space of the warehouse to collect environmental parameters, real-time location and operational status information of physical entities.

3. The intelligent warehousing and logistics sorting system according to claim 2, characterized in that, The distributed sensor network further includes: Temperature and humidity sensor arrays are installed on pre-set grid nodes in the warehouse, with the temperature sensor having a measurement accuracy of no less than ±0.3℃ and the humidity sensor having a measurement accuracy of no less than ±3%RH. The ultra-wideband positioning system consists of positioning base stations and positioning tags. It is configured to provide the three-dimensional spatial coordinates of the positioning tags attached to automated transport equipment, containers or operators at an update frequency of not less than 15 Hz, with a positioning error of not more than 20 centimeters. An ultrasonic ranging sensor, embedded inside a transport container, is designed to calculate the container's space utilization rate in real time by measuring the height distribution of materials inside the container.

4. The intelligent warehousing and logistics sorting system according to claim 1, characterized in that, The three-dimensional simulation model constructed and maintained by the global state awareness and simulation model construction module includes: A dynamic weighted warehouse topology model, where nodes represent functional areas within the warehouse, edges represent connecting paths between functional areas, and the weights of the edges are adjusted in real time based on the dynamic flow coefficients calculated from the real-time density and average speed of automated transport equipment on the path. The full resource status matrix is ​​set to record the status of each automated transportation device and each transportation container in real time; The material lifecycle state vector is set to record the precise storage location, quantity, batch information, and cumulative environmental impact value of each material in the inventory, which is obtained by continuous integration calculation based on the real-time environmental data of its location and exposure duration.

5. The intelligent warehousing and logistics sorting system according to claim 4, characterized in that, The dynamically weighted warehouse topology model is stored in a relational database; the nodes represent functional areas such as storage locations, charging stations, operating tables, or inbound / outbound channels; the basic weight of the edges is the physical distance; and the update cycle of the dynamic flow coefficient is no more than 3 seconds.

6. The intelligent warehousing and logistics sorting system according to claim 4, characterized in that, The three-dimensional simulation model also includes a three-dimensional environmental distribution map generated by interpolation algorithm from the data collected by the temperature and humidity sensor array. This map represents the real-time temperature and humidity status of any coordinate point in the warehouse in the form of a voxel grid.

7. The intelligent warehousing and logistics sorting system according to claim 1, characterized in that, The demand forecasting and analysis module employs a deep learning model based on convolutional neural networks. The input layer of the model receives historical order demand time series data in terms of material categories, along with corresponding external variables such as market activities and seasonal factors. The model takes the actual demand data of the previous period and the production plan for the future preset period as input, and outputs a probabilistic demand forecast vector. This vector gives the probability and expected quantity of each material being demanded in each time window within the future preset period, and generates a demand distribution map indicating the future picking pressure in different areas of the warehouse.

8. The intelligent warehousing and logistics sorting system according to claim 1, characterized in that, The task optimization and resource allocation module dynamically partitions orders by running a multi-objective particle swarm optimization algorithm. The multi-objective particle swarm optimization algorithm evaluates each order partitioning scheme using a preset multi-objective fitness function. The fitness function is a weighted linear combination function containing at least three of the following terms: A path cost function designed to minimize the total length of planned paths for all task groups; A container efficiency function designed to maximize the space utilization of all containers used. The aim is to minimize the time delay function that exceeds the expected completion time of an order from its committed service level agreement; An impact function designed to minimize the incremental environmental impact of sensitive materials during picking and transportation.

9. The intelligent warehousing and logistics sorting system according to claim 8, characterized in that, The task optimization and resource allocation module further includes a resource allocation and path planning unit, which is configured to select the final execution scheme from a set of optimal solutions found by the multi-objective particle swarm optimization algorithm, and select suitable automated transportation equipment and transportation containers for each task group in the scheme from the full resource state matrix according to the total weight and total volume of materials in the group, and call the improved A* path search algorithm to calculate a real-time optimal path for the selected automated transportation equipment on the dynamic weight warehouse topology model.

10. An intelligent warehousing and logistics sorting method, characterized in that, Includes the following steps: The multi-dimensional data acquisition module acquires and integrates dynamic and static information related to warehousing operations from multiple heterogeneous sources; The global state perception and simulation model building module receives data provided by the multi-dimensional data acquisition module and builds and maintains a three-dimensional simulation model that is synchronized with the physical warehouse in real time. The demand forecasting and analysis module uses historical business data and future planning information to probabilistically estimate the outbound demand for materials and generate a future demand distribution map. The task optimization and resource allocation module receives new outbound orders and, in conjunction with the real-time status information of the 3D simulation model and the future demand distribution map, dynamically divides the outbound orders into several task groups and allocates suitable execution resources and optimal paths to each task group. The execution control and feedback adjustment module transforms the instructions generated by the task optimization and resource configuration module into specific operation commands for the execution device, and performs real-time monitoring and dynamic adjustment of the execution process.

Images