Warehouse management system and management method

Abstract

The invention provides a warehouse management system and management method, and relates to the field of warehouse management. Comprising a storage intelligent twin system and a storage self-organization execution system, and the storage intelligent twin system is used for collecting storage environment data and generating a storage scheduling scheme based on intelligent analysis and calculation optimization. The storage self-organization execution system executes storage adjustment, facility self-repairing, energy consumption optimization management and cargo full life cycle tracking according to the optimization scheme; the system adopts a closed-loop operation mode of multi-dimensional data perception, storage simulation, behavior optimization and global dynamic feedback to realize collaborative optimization of data acquisition, intelligent decision, execution control and real-time adjustment; the flexible form storage unit, the material self-adaptive storage rack, the self-repairing storage facility and the energy management unit can be dynamically adjusted according to storage requirements, and the storage efficiency and the system stability are improved.

Description

Technical Field

[0001] The present invention relates to the field of warehouse management, and specifically to a warehouse management system and a management method. Background Art

[0002] In the fields of modern logistics and supply chain management, the warehouse system undertakes the core tasks of goods storage, scheduling, and distribution; with the rapid development of e-commerce and manufacturing, warehouse management faces problems such as an increasing variety of goods, an increasing frequency of inbound and outbound, and limited storage space; traditional warehouse management relies on manual operations, with low efficiency and high costs, and it is difficult to meet the current market's demand for rapid response and precise management.

[0003] Existing intelligent warehouse management technologies mainly rely on the Internet of Things, RFID identification, and automated equipment. The Internet of Things technology is used to monitor the status of goods in real time, RFID tags achieve automatic identification of goods, and automated equipment performs storage and picking operations through robotic arms or intelligent robots; some of these systems use digital twin technology to build a virtual model of the warehouse to optimize the placement of goods and the picking path; these solutions have improved warehouse efficiency to a certain extent, but there are still many challenges in the face of complex warehouse environments and high-frequency dynamic scheduling.

[0004] Existing technologies have deficiencies in multi-dimensional data processing, intelligent scheduling optimization, and long-term stability of warehouse facilities; many systems lack efficient data fusion capabilities, resulting in inaccurate perception of the warehouse environment, which in turn affects the optimization of storage and scheduling strategies; on the other hand, existing technologies have limitations in goods tracking and energy consumption management and are difficult to meet the needs of long-term operation; in addition, warehouse equipment may experience wear and failures during long-term use, and existing systems still have defects in facility self-repair and structural optimization. Summary of the Invention

[0005] (1) Technical Problems to be Solved

[0006] In view of the deficiencies of the existing technologies, the present invention provides a warehouse management system and a management method to solve the problems of inaccurate perception of the warehouse environment, insufficient optimization ability of storage and scheduling strategies, limited goods tracking, inefficient energy consumption management, and lack of maintenance and repair capabilities during the long-term use of warehouse equipment.

[0007] (2) Technical Solutions

[0008] To achieve the above objectives, the present invention is realized through the following technical solutions: A warehouse management system and a management method, including a warehouse intelligent twin system and a warehouse self-organizing execution system;

[0009] The described intelligent warehouse digital twin system includes a multi-dimensional data perception layer, a storage simulation layer, a behavior optimization layer, and a global dynamic feedback layer. The multi-dimensional data perception layer collects warehouse environment data and transmits it to the storage simulation layer for analysis. The storage simulation layer generates a warehouse scheduling plan based on computational optimization and transmits the optimization results to the behavior optimization layer. The behavior optimization layer calculates the optimal execution strategy and issues it to the warehouse self-organizing execution system. The global dynamic feedback layer monitors the execution status and transmits the feedback data to the storage simulation layer and the behavior optimization layer. The warehouse self-organizing execution system includes a flexible form storage unit, a material adaptive storage rack, a self-repairing warehouse facility, an energy management unit, and a non-contact full-life-cycle goods tracking unit, which are used to make dynamic adjustments according to the instructions of the behavior optimization layer.

[0010] Preferably, the multi-dimensional data perception layer includes a hyperspectral vision sensor, a nanoscale stress sensing unit, a light field fluid perception module, and a multi-modal storage state detection unit. The hyperspectral vision sensor is used to perform spectral scanning on the surface features of goods in the storage environment to obtain comprehensive data on changes in the surface composition of goods, color deviation, surface contamination, and the influence of temperature and humidity. The spectral scanning technology can accurately identify possible pollutants and color differences on the surface of goods. The hyperspectral vision sensor adopts a distributed spectral sensing unit, where each spectral sensing unit consists of a grating beam splitting module, a photoelectric detection array, an adjustable filter group, and a data conversion chip. The grating beam splitting module separates light of different bands, and the adjustable filter group screens spectral bands. Finally, the photoelectric detection array collects the light intensity signal, and the data conversion chip performs analog-to-digital conversion to store the spectral data in the local cache for further analysis and processing. The nanoscale stress sensing unit is based on a micro-machined strain sensor and a piezoresistive strain gauge. Through nanoscale stress sensors attached to key stressed parts of the storage rack, it detects changes in stress data such as bending stress, shear stress, and dynamic impact load of the storage rack. When the storage rack is subjected to external pressure, impact, or uneven load, the nanoscale stress sensing unit can promptly sense and record these changes. At the same time, the stress data is amplified, filtered, and denoised by the MEMS signal processing chip and then transmitted to the central data processing unit for further analysis. The light field fluid perception module is based on interferometric fluid measurement technology, which includes a micro thermal anemometer, a flow field reconstruction algorithm module, and a data fusion interface, and is used to analyze changes in air velocity, flow direction, and pressure in the storage area. The cooperation of these components can accurately measure small changes in air flow in the storage environment, thereby helping to optimize air circulation and temperature and humidity control. The multi-modal storage state detection unit uses an infrared thermal imaging sensor, a millimeter-wave scanner, and a VOC gas sensor to comprehensively monitor the storage state of goods. The infrared thermal imaging sensor is used to detect temperature changes on the surface of goods. The millimeter-wave scanner measures the density and loss of goods through multi-frequency millimeter-wave reflection. The VOC gas sensor detects the volatile organic gas situation in the goods storage environment through a semiconductor gas detection element. All data is transmitted to the storage simulation layer through low-power wireless communication technology for further calculation.

[0011] Preferably, the storage simulation layer is based on a high-dimensional warehousing mapping model and a high-dimensional computing optimization engine. The high-dimensional warehousing mapping model uses dynamic topology analysis technology and a warehousing topology optimization algorithm. The dynamic topology analysis technology is based on a hierarchical Delaunay triangulation algorithm. By hierarchically extracting key storage nodes from the warehousing space and combining actual cargo flow data, a high-dimensional storage space model is formed, providing data support for subsequent warehousing optimization. The warehousing topology optimization algorithm combines Voronoi partitioning technology to calculate the optimal storage area distribution according to the characteristics and requirements of the goods, and finally generates a dataset for subsequent optimization. The high-dimensional computing optimization engine uses a reinforcement learning model (DQN) and intelligent evolutionary computing. The intelligent evolutionary computing is a genetic algorithm GA. By analyzing the warehousing historical data and real-time storage status, the warehousing layout and storage strategy are optimized. The DQN model is modeled through a state space, an action space, and a reward function. The warehousing state is modeled as a multi-dimensional state vector, and the warehousing adjustment action is used as an optional optimization plan. The optimal scheduling strategy is calculated based on an evaluation function of warehousing efficiency. The genetic algorithm GA uses a fitness function selection strategy to optimize the storage strategy through crossover and mutation, and combines the Bayesian optimization algorithm to dynamically adjust the storage unit distribution. Finally, the calculated optimized storage path and access strategy will be transmitted to the behavior optimization layer for further execution.

[0012] Preferably, the behavior optimization layer is based on warehousing objectives, storage distribution, picking paths, and energy consumption control strategies, and is used to calculate the warehousing execution plan and make dynamic adjustments based on real-time feedback. The behavior optimization layer uses a Markov decision process (MDP) and a multi-objective optimization algorithm. The Markov decision process is based on a state transition matrix. By calculating the optimal action strategy in different warehousing states, the adaptive optimization of cargo storage is realized. The multi-objective optimization algorithm uses a non-dominated sorting genetic algorithm (NSGA-II). Optimization objectives are constructed based on factors such as storage density, picking efficiency, cargo safety, and energy consumption cost, and the optimal warehousing execution strategy is calculated using the Pareto optimal solution screening method. The final scheduling decision of the behavior optimization layer will be transmitted to the warehousing self-organizing execution system and dynamically adjusted in combination with the data of the global dynamic feedback layer.

[0013] Preferably, the global dynamic feedback layer adopts an adaptive monitoring unit, a data analysis unit and an intelligent control module. The adaptive monitoring unit includes a laser displacement sensor, an inertial measurement unit (IMU), and a low-power vibration sensor, which are used to monitor the real-time state of the storage rack and goods and transmit the data to the data analysis unit. The data analysis unit uses a real-time data fusion algorithm, extracts effective data through wavelet transform noise reduction technology, and predicts the change trend of the storage state based on the adaptive Kalman filter algorithm. The intelligent control module dynamically adjusts the storage execution unit through a fuzzy control algorithm and a model predictive control strategy (MPC), combined with the warehousing optimization parameters provided by the storage simulation layer, to ensure the adaptability of the storage strategy.

[0014] Preferably, the flexible form storage unit adopts a liquid metal skeleton and a dynamic tension adjustment structure. The liquid metal skeleton is composed of a gallium-indium alloy and a nano-composite polymer. In the low-power state, it can change the support form of the shelf through an electromagnetic induction control unit under the condition of being lower than the energy consumption threshold set by the system. The threshold is the maximum power that can be consumed during operation. This energy consumption threshold is preset according to the overall energy consumption budget and operation efficiency requirements of the warehousing system to ensure that while optimizing the storage space, the efficient use of energy is also taken into account to adapt to the storage needs of different goods. The dynamic tension adjustment structure adopts a shape memory polymer and a local thermal drive unit, and adjusts the rigidity and support capacity of the storage unit through current control to ensure the best support effect when carrying different goods; and combines the data of the global dynamic feedback layer to optimize the deformation mode of the storage rack in real time to ensure the maximum utilization of the warehousing space.

[0015] Preferably, the material adaptive storage rack adopts shape memory alloy and programmable micro-nano mechanical structure. The shape memory alloy is composed of nickel-titanium alloy (NiTi) and copper-aluminum-nickel alloy (Cu-Al-Ni), and has the characteristic of reversible deformation. It can change its shape under specific temperature stimulation to adapt to the storage needs of different goods. The storage rack is in the standard shelf structure in the initial state. When receiving the storage adjustment instruction from the behavior optimization layer, the shape memory alloy is heated to the phase transition temperature through the electrothermal drive module, causing local deformation of the storage rack structure to adjust the storage space size and the shelf arrangement mode. The programmable micro-nano mechanical structure consists of a micro-electro-mechanical system (MEMS) unit, a nano-level servo controller, and an adaptive stiffness adjustment unit. The micro-electro-mechanical system (MEMS) unit controls the adjustable support structure of the storage rack through a piezoelectric drive device to ensure the best rigidity and flexibility matching of the shelf when carrying different weights of goods. The nano-level servo controller, based on the closed-loop feedback control algorithm and combined with the data of the global dynamic feedback layer, adjusts the deformation mode and support strength of the storage rack when detecting abnormal force on the storage rack or mismatched cargo shape to ensure the safety and stability of the goods during storage. The adaptive stiffness adjustment unit is composed of carbon nanotube reinforced composite materials, which can adjust the stiffness of the storage rack under different load states and optimize the load-bearing capacity and morphological stability of the storage rack by adjusting the arrangement direction of carbon nanotubes through the electric field excitation control module. The material adaptive storage rack combines the optimal storage state of the goods calculated by the storage simulation layer and the dynamic adjustment parameters provided by the behavior optimization layer to achieve the intelligent adaptation and long-term stability of the storage rack.

[0016] Preferably, the self-healing storage facility adopts a material stress sensing device, a local self-healing coating, a microcapsule repair unit and an intelligent repair management unit. The material stress sensing device consists of a fiber Bragg grating sensor (FBG), an ultrasonic crack detection unit and a stress analysis processor. The fiber Bragg grating sensor is embedded in the key load-bearing parts of the storage rack to monitor the stress distribution in real time. When the storage rack undergoes structural fatigue or the stress exceeds the safe range, the stress analysis processor triggers the repair mechanism through the stress threshold warning algorithm. The ultrasonic crack detection unit is based on high-frequency ultrasonic scanning technology. It uses piezoelectric transducers to detect cracks and deformations on the surface and inside of the storage rack, and calculates the crack propagation trend through non-linear acoustic wave analysis methods to ensure that the repair system can take active compensation measures before damage occurs. The local self-healing coating consists of polymer microcapsules, nano-catalysts and liquid metal repair agents. When cracks or wear appear on the surface of the storage rack, the microcapsules release the repair agent through the diffusion of micro-cracks and quickly solidify and fill the cracks under the action of nano-catalysts to ensure the integrity and durability of the storage rack surface. The microcapsule repair unit releases the repair agent in the microcapsules through a temperature control activation mechanism when abnormal temperature changes or uneven stress distribution of the storage rack are detected, and combines the monitoring data of the global dynamic feedback layer to realize the intelligent self-healing function of the storage rack. The intelligent repair management unit combines machine learning prediction algorithms to predict the fatigue life based on the historical data, environmental conditions and load status of the storage rack, and actively triggers the repair process when the storage rack enters the high-loss stage to ensure the long-term stable operation of the storage facility.

[0017] Preferably, the energy management unit includes a phase change energy storage unit, an intelligent air flow control unit, and a local heat exchange network. The phase change energy storage unit is based on high-efficiency thermoelectric conversion materials, microcapsule phase change energy storage materials, and nanoscale temperature control coatings. Among them, the high-efficiency thermoelectric conversion materials use Bi2Te3-based thermoelectric semiconductors. Combining the temperature monitoring data provided by the global dynamic feedback layer, local temperature control is carried out through the thermoelectric coupling transmission mechanism. When the temperature in the storage area rises abnormally, the thermoelectric conversion materials convert thermal energy into electrical energy through a thermocouple array and store the electrical energy in the battery management system. When the temperature drops, the stored energy is released through the reverse conduction mechanism to optimize the temperature stability of the storage environment. The intelligent air flow control unit uses a fluid bionic air duct structure, an intelligent variable wind speed fan, and a nanoscale low-resistance filter screen. Among them, the fluid bionic air duct structure optimizes the internal air flow path in the warehouse based on the turbulent flow control algorithm to reduce energy loss. The intelligent variable wind speed fan uses an EC DC variable frequency motor and combines the PID dynamic wind speed control algorithm to automatically adjust the wind speed according to the temperature and humidity conditions in the storage area to maintain the best storage environment with the minimum energy consumption. The local heat exchange network uses a solid-state heat conduction module, a microfluidic cooling system, and a heat transfer optimization controller. Among them, the solid-state heat conduction module performs local heat exchange near the storage rack through a thermopile array to prevent local temperature fluctuations from affecting the storage environment. The microfluidic cooling system is based on a microcapillary structure and an intelligent fluid control valve, and improves the cooling efficiency and reduces the system energy consumption through adaptive fluid flow rate adjustment. The heat transfer optimization controller combines the monitoring data of the global dynamic feedback layer to adjust the energy consumption distribution in the storage environment in real time to ensure the energy-saving operation of the storage system.

[0018] Preferably, the non-contact full-life-cycle goods tracking unit uses quantum dot labeling and a spectral scanning recognition system. The quantum dot labeling technology is based on CdSe / ZnS core-shell structure quantum dot materials, and forms a unique spectral fingerprint identification on the surface of the goods through spraying, nano-printing, or molecular coating technology. The spectral scanning recognition system uses a Fourier transform infrared spectrometer (FTIR) and a multi-band fluorescence imaging system. During the flow of goods, the FTIR system is used to quickly match the spectral characteristics of the goods, and combined with the deep learning spectral recognition algorithm to ensure the accuracy of goods recognition. The non-contact full-life-cycle goods tracking unit is data-linked with the global dynamic feedback layer. During the flow of goods, the storage status of the goods is monitored in real time, and dynamic adjustment is carried out in combination with the storage plan provided by the behavior optimization layer to ensure the safety, stability, and traceability of the goods during storage.

[0019] (III) Beneficial effects

[0020] The present invention provides a warehouse management system and a management method. It has the following beneficial effects:

[0021] 1. Through the deep integration of the warehousing intelligent twin system and the warehousing self-organizing execution system, combined with multi-dimensional data perception, storage simulation calculation, behavior optimization scheduling, and global dynamic feedback, the present invention realizes the adaptive optimization and efficient management of intelligent warehousing; the system adopts a high-dimensional computing optimization engine, dynamically adjusts the storage distribution, material flow direction, and access path based on real-time data, can reduce the storage redundancy by more than 30%, improve the access efficiency by more than 40%, and optimizes the warehousing temperature control and energy consumption distribution through the energy management unit, reducing the overall operation cost by more than 20% and ensuring efficient and energy-saving operation.

[0022] 2. The present invention adopts self-repairing warehousing facilities and non-contact full-life-cycle goods tracking technology. Through the material stress sensing device, local self-repairing coating, and self-adaptive micro-growth material, combined with the predictive maintenance algorithm, the warehousing structure has the capabilities of automatic detection, early warning, and repair, can reduce the maintenance cost by more than 30%, and increase the equipment life by more than 3 times; at the same time, the system integrates quantum dot marking and spectral scanning recognition to ensure the full-life-cycle tracking of goods from warehousing, storage, retrieval, to transportation, realizes precise management, and improves the warehousing safety and traceability. Detailed implementation manners

[0023] The technical solutions in the embodiments of the present invention will be clearly and completely described below. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0024] Embodiment 1:

[0025] An embodiment of the present invention provides a warehousing management system and a management method. Specifically, in a large logistics warehouse, the warehousing management system and the management method are introduced. First, a multi-dimensional data perception layer is deployed, and hyperspectral vision sensors are installed and distributed at various key positions in the warehouse, such as the goods entrance, storage area passages, etc. The distributed spectral perception unit of the sensor starts to work. The grating beam splitting module separates light of different bands, the tunable filter group screens spectral bands, the photoelectric detection array collects light intensity signals, and the data conversion chip performs analog-to-digital conversion and stores the spectral data in the local cache. At the same time, the nano-level stress sensing unit is attached to the key stressed parts of the storage rack to detect changes in stress data of bending stress, shear stress, and dynamic impact load. After being amplified, filtered, and denoised by the MEMS signal processing chip, it is transmitted to the central data processing unit. The micro thermal anemometer, flow field reconstruction algorithm module, and data fusion interface of the optical field fluid perception module cooperate to analyze the air velocity, flow direction, and pressure changes in the warehousing area. The infrared thermal imaging sensor, millimeter wave scanner, and VOC gas sensor of the multi-modal storage state detection unit comprehensively monitor the storage state of the goods, and all data are transmitted to the storage simulation layer through low-power wireless communication technology.

[0026] Then the storage simulation layer works based on the high-dimensional warehousing mapping model and the high-dimensional computing optimization engine. The high-dimensional warehousing mapping model uses dynamic topology analysis technology and warehousing topology optimization algorithms. The dynamic topology analysis technology extracts key storage nodes using the hierarchical De launay triangulation algorithm and forms a high-dimensional storage space model in combination with the goods circulation data. The warehousing topology optimization algorithm calculates the optimal storage area distribution in combination with the Voronoi partitioning technology and generates a data set. The high-dimensional computing optimization engine adopts the DQN model and the genetic algorithm GA. The DQN model is modeled through the state space, action space, and reward function to calculate the optimal scheduling strategy. The genetic algorithm GA selects strategies using the fitness function, optimizes the storage strategy, and dynamically adjusts the storage unit distribution in combination with the Bayesian optimization algorithm. Finally, the optimization result is transmitted to the behavior optimization layer.

[0027] Immediately afterwards, the behavior optimization layer calculates the warehousing execution plan based on the warehousing objectives, storage distribution, picking path, and energy consumption control strategy, using the MDP and NSGA-II algorithms. The MDP calculates the optimal action strategy based on the state transition matrix, and the NSGA-II constructs an optimization objective based on multiple factors and screens the Pareto optimal solutions. After obtaining the optimal warehousing execution strategy, it is transmitted to the warehousing self-organization execution system and dynamically adjusted in combination with the data of the global dynamic feedback layer.

[0028] Then, the adaptive monitoring unit, data analysis unit, and intelligent control module of the global dynamic feedback layer work together. The laser displacement sensor, IMU, and low-power vibration sensor of the adaptive monitoring unit monitor the real-time status of the storage rack and goods, and transmit it to the data analysis unit. The data analysis unit uses real-time data fusion algorithms, wavelet transform noise reduction techniques, and adaptive Kalman filter algorithms to extract effective data and predict the changing trend of the storage status. The intelligent control module dynamically adjusts the storage execution unit through fuzzy control algorithms and MPC, combined with the warehousing optimization parameters provided by the storage simulation layer.

[0029] Among them, the flexible form storage unit in the warehousing self-organizing execution system adopts a liquid metal skeleton and a dynamic tension adjustment structure. In the low-power state, the electromagnetic induction control unit changes the support form of the shelf. The dynamic tension adjustment structure adjusts the rigidity and support capacity of the storage unit through current control, and combines the data of the global dynamic feedback layer to optimize the deformation mode of the storage rack; the material adaptive storage rack adopts shape memory alloy and programmable micro-nano mechanical structure. The shape memory alloy deforms by heating through the electrothermal drive module. The programmable micro-nano mechanical structure adjusts the deformation mode and support strength of the storage rack through piezoelectric drive devices, nano-level servo controllers, and adaptive stiffness adjustment units to achieve intelligent adaptation and long-term stability; the self-repairing warehousing facilities adopt material stress sensing devices, local self-repairing coatings, microcapsule repair units, and intelligent repair management units. The material stress sensing device monitors the stress distribution and triggers the repair mechanism. The local self-repairing coating and microcapsule repair unit release repair agents to fill the cracks. The intelligent repair management unit actively triggers the repair process based on machine learning prediction algorithms; the energy management unit includes a phase change energy storage unit, an intelligent air flow control unit, and a local heat exchange network. The phase change energy storage unit performs local temperature control through thermoelectric conversion materials. The intelligent air flow control unit optimizes the air flow path and automatically adjusts the wind speed. The local heat exchange network conducts local heat exchange and improves the cooling efficiency to ensure energy-saving operation; the non-contact full-life-cycle goods tracking unit adopts quantum dot marking and spectral scanning recognition systems. The quantum dot marking technology forms a spectral fingerprint identification on the surface of the goods. The spectral scanning recognition system ensures the accuracy of goods identification through the FTIR system and deep learning spectral recognition algorithms, and is linked with the data of the global dynamic feedback layer to monitor the real-time status of goods storage and dynamically adjust it to ensure the safety, stability, and traceability of goods storage.

[0030] Embodiment 2:

[0031] The differences between this embodiment and Embodiment 1 are as follows: This embodiment applies the warehousing management system and management method in a small e-commerce warehouse. In the deployment of the multi-dimensional data perception layer, considering the limited space and relatively concentrated types of goods in a small warehouse, the quantity and distribution positions of hyperspectral vision sensors are optimized and adjusted. They are mainly installed in the areas where goods are intensively stored and at the entrances and exits of frequently accessed channels to more accurately obtain the surface feature data of goods at key positions; in addition to attaching to the key stressed parts of the storage racks, the nano-level stress sensing unit also adds the monitoring of the connection parts of small shelves because the structure of small shelves is relatively simple, and the stress changes at the connection parts have a greater impact on the overall stability; the light field fluid perception module in a small warehouse focuses more on monitoring the small airflow changes in local warehousing areas to quickly adjust the ventilation equipment and maintain a suitable storage environment; the monitoring frequencies of the infrared thermal imaging sensor, millimeter wave scanner, and VOC gas sensor in the multi-modal storage state detection unit are increased to adapt to the characteristics of fast turnover and frequent changes in storage states of goods in a small warehouse, ensuring timely acquisition of accurate storage state data.

[0032] Among them, when constructing the high-dimensional warehousing mapping model in the storage simulation layer, according to the characteristics of the warehousing space layout of a small warehouse, a more refined method of extracting key storage nodes layer by layer is adopted, combined with the characteristics of e-commerce goods such as size, weight, and sales frequency, to form a more practical high-dimensional storage space model; when calculating the optimal storage area distribution, the warehousing topology optimization algorithm fully considers the volatility of goods storage requirements during e-commerce promotion activities and dynamically adjusts the storage area division to improve the flexibility and utilization rate of the warehousing space; when optimizing the warehousing layout and storage strategy, the high-dimensional computing optimization engine pays more attention to the control of energy consumption costs. Through the optimization of intelligent evolutionary computing and DQN model, a low-energy consumption and high-efficiency warehousing solution is customized for small warehouses.

[0033] Immediately afterwards, when calculating the warehousing execution plan in the behavior optimization layer, combined with the warehousing objectives of small warehouses and the characteristics of e-commerce operations, such as fast shipment and improved customer satisfaction, the picking path and energy consumption control strategy are optimized specifically. More flexible MDP and NSGA-II algorithms are adopted to quickly respond to changes in the warehousing state and adjust the warehousing execution strategy in real time to meet the strict requirements of e-commerce warehouses for warehousing efficiency and cost control.

[0034] The application of the global dynamic feedback layer in small warehouses pays more attention to real-time performance and accuracy. The sensors of the adaptive monitoring unit have higher precision and can capture the minute state changes of storage racks and goods in real time. The data analysis unit adopts more advanced data fusion algorithms and prediction models to quickly and accurately analyze and predict the trend of storage state changes. The intelligent control module combines the warehousing optimization parameters of small warehouses and realizes precise dynamic adjustment of the storage execution unit through more refined fuzzy control algorithms and MPC, ensuring the adaptability and efficiency of storage strategies.

[0035] Among them, the implementation of the warehousing self-organizing execution system in small warehouses has also been adjusted. The liquid metal skeleton and dynamic tension adjustment structure of the flexible form storage unit are more lightweight and flexible, capable of quickly adapting to the storage needs of e-commerce goods of different sizes while reducing energy consumption. When the shape memory alloy and programmable micro-nano mechanical structure of the material adaptive storage rack adjust the storage space size and shelf arrangement, more attention is paid to space utilization and the convenience of goods access. The self-repairing warehousing facilities in small warehouses focus more on preventive maintenance, using material stress sensing devices to give early warnings of potential structural damage and promptly activating the repair mechanism to reduce warehousing interruptions caused by facility failures. The energy management unit in small warehouses realizes the minimization of energy consumption and reduces operating costs through more precise phase change energy storage units, intelligent air flow regulation units, and local heat exchange network control. The non-contact full-life-cycle goods tracking unit in small warehouses is deeply integrated with the e-commerce order management system to update the goods storage status and order information in real time, improving customer satisfaction and warehouse operation efficiency.

[0036] Although the embodiments of the present invention have been shown and described, those of ordinary skill in the art can understand that various changes, modifications, substitutions, and variations can be made to these embodiments without departing from the principles and spirit of the present invention. The scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A warehouse management system, including a warehouse intelligent twin system and a warehouse self-organizing execution system; The warehouse intelligent twin system includes a multidimensional data perception layer, a storage simulation layer, a behavior optimization layer and a global dynamic feedback layer. The multidimensional data perception layer collects warehouse environment data and transmits it to the storage simulation layer for analysis. The storage simulation layer generates a warehouse scheduling plan based on computational optimization and transmits the optimization results to the behavior optimization layer. The behavior optimization layer calculates the optimal execution strategy and sends it to the warehouse self-organizing execution system. The global dynamic feedback layer monitors the execution status and transmits feedback data to the storage simulation layer and the behavior optimization layer. The warehouse self-organizing execution system includes a flexible morphological storage unit, a material adaptive storage rack, a self-repairing storage facility, an energy management unit and a non-contact full life cycle cargo tracking unit, which is used to make dynamic adjustments according to the instructions of the behavior optimization layer.

2. A warehouse management system according to claim 1, characterized in that: The multi-dimensional data perception layer includes a hyperspectral vision sensor, a nano-scale stress sensing unit, a light field fluid perception module and a multi-modal storage status detection unit. The hyper-spectral vision sensor is used to detect the spectral characteristics, composition and temperature and humidity changes of the goods. The nano-scale stress sensing unit is used to monitor the load state, structural deformation and pressure distribution of the storage rack. The light field fluid perception module is used to analyze the air flow, energy loss and micro-environment stability inside the warehouse. The multi-modal storage status detection unit is used to analyze material aging, weight loss and pollution risks during the storage of goods. The multi-dimensional data perception layer transmits the data to the storage simulation layer for computational analysis, and transmits the analysis results to the behavior optimization layer for decision optimization.

3. A warehouse management system according to claim 1, characterized in that: The storage simulation layer is based on a high-dimensional warehouse mapping model and a high-dimensional computing optimization engine, combined with a machine learning prediction algorithm, to generate a real-time warehouse distribution model, a cargo flow optimization strategy, and an environmental impact assessment. The high-dimensional warehouse mapping model uses dynamic topology analysis technology and a warehouse topology optimization algorithm, and can autonomously learn historical storage patterns and optimization strategies, and adjust the cargo storage structure based on the data provided by the multidimensional data perception layer. The high-dimensional computing optimization engine calculates the optimal path distribution of cargo at different storage locations based on topology analysis and an adaptive scheduling matrix, and makes real-time adjustments in combination with the warehouse status information provided by the global dynamic feedback layer, and transmits the optimization results to the behavior optimization layer so that the execution system can take corresponding adjustments.

4. A warehouse management system according to claim 1, characterized in that: The behavior optimization layer calculates the storage execution plan based on the storage target, storage distribution, picking path, and energy consumption control strategy. The behavior optimization layer transmits the calculated storage execution plan to the global dynamic feedback layer in real time. The global dynamic feedback layer evaluates and corrects the optimization results of the behavior optimization layer according to the monitored operating status of the storage self-organizing execution system. If deviations or abnormalities are found during the execution process, timely feedback is given to the behavior optimization layer so that it can readjust the optimization strategy. The global dynamic feedback layer is used to monitor the operating status of the warehouse self-organizing execution system, and adjust the storage and scheduling strategies in combination with the optimization results of the behavior optimization layer. The global dynamic feedback layer generates feedback data by real-time monitoring of storage unit status, warehouse environment changes, and equipment operation conditions, and transmits the feedback data to the storage simulation layer and the behavior optimization layer, thereby realizing dynamic optimization of the warehouse system and an intelligent decision-making closed loop.

5. A warehouse management system according to claim 1, characterized in that: The flexible morphological storage unit is based on a liquid metal skeleton and a dynamic tension adjustment structure. The liquid metal skeleton can change the supporting shape of the shelf under low energy consumption conditions, so that the storage space can adapt to changes in the size and shape of the goods. The dynamic tension adjustment structure is based on shape memory polymers and local thermal drive units. When storage needs change, the shelf structure is adjusted to reduce waste of storage space, improve storage efficiency, and adjust the morphological parameters according to the decision results of the behavior optimization layer.

6. A warehouse management system according to claim 1, characterized in that: The material adaptive storage rack is based on shape memory alloy and programmable micro-nano mechanical structure, which can adjust the rigidity, flexibility and shape of the storage rack according to the storage requirements of goods. The shape memory alloy adopts nickel-titanium alloy and copper-aluminum-nickel alloy, and has reversible deformation characteristics. After receiving the storage adjustment instruction, the temperature is raised to the phase change temperature through the electric thermal drive module to make the storage rack structure partially deformed, so as to realize the dynamic adjustment of the space size and shelf arrangement; the programmable micro-nano mechanical structure is composed of a micro-electromechanical system (MEMS) unit and a nano-level servo controller, which ensures that the storage rack has the best supporting capacity under different loads, and optimizes and adjusts the supporting structure in real time according to the data of the global dynamic feedback layer to ensure the safety and stability of the goods.

7. A warehouse management system according to claim 1, characterized in that: The self-repairing storage facility is based on a material stress sensing device, a local self-repairing coating, a microcapsule repair unit and an intelligent repair management unit. The material stress sensing device adopts a nano-scale strain sensor to monitor the structural fatigue of the storage rack and the storage facility in real time. The local self-repairing coating adopts microcapsule encapsulation technology to automatically release the repair agent when the material is damaged. The microcapsule repair unit releases the repair agent in the microcapsule by detecting the temperature change of the storage rack, and combines the monitoring data of the global dynamic feedback layer to realize the intelligent self-repair function of the storage rack. The intelligent repair management unit combines a machine learning prediction algorithm to predict fatigue life based on the storage rack historical data, environmental conditions, and load status, and actively triggers the repair process when the storage rack enters the high-loss stage to ensure the long-term stable operation of the storage facility.

8. A warehouse management system according to claim 1, characterized in that: The energy management unit is based on a phase change energy storage unit, an intelligent airflow control unit and a local heat exchange network. The phase change energy storage unit adopts high-efficiency thermoelectric conversion materials to automatically adjust the storage temperature when the ambient temperature changes. The intelligent airflow control unit is based on a fluid bionic structure to optimize the air flow in the warehouse to reduce energy consumption losses. The local heat exchange network adopts solid-state heat conduction technology to locally adjust the temperature in the storage area to improve energy utilization, and optimizes energy consumption distribution in combination with the decision-making plan of the behavior optimization layer.

9. A warehouse management system according to claim 1, characterized in that: The contactless full-life cycle cargo tracking unit is based on quantum dot labeling and spectral scanning recognition. The quantum dot labeling technology is used to implant a unique and tamper-proof identity code in the cargo production stage. The spectral scanning recognition unit adopts adaptive filtering technology to monitor the cargo identity in real time during the warehouse management process, and optimizes the cargo storage plan in combination with the storage environment data to ensure the safety and traceability of the entire warehousing process, and performs dynamic adjustments in combination with the global dynamic feedback layer.

10. A warehouse management method according to claim 1, characterized in that: Collecting storage environment information through the multi-dimensional data perception layer, and transmitting the data to the storage simulation layer for calculation and analysis; The storage simulation layer generates optimal storage and cargo flow solutions in combination with machine learning predictions, and transmits the optimization results to the behavior optimization layer; The behavior optimization layer issues execution instructions to the flexible form storage unit, the material adaptive storage rack, the self-repairing storage facility, and the energy management unit; The global dynamic feedback layer monitors the operating status of the storage system and provides feedback data to the storage simulation layer and the behavior optimization layer; The storage simulation layer adjusts optimization parameters according to the data provided by the global dynamic feedback layer to achieve intelligent scheduling and adaptive optimization of the storage system.