Digital twin system and method for visual warehousing based on big data analysis

Abstract

The present invention discloses a digital twin system and method for visualized warehousing based on big data analysis. The present invention relates to the field of warehousing, and includes a data acquisition unit, a data transmission unit, a data processing unit, a data analysis unit, and a digital twin model unit. The data acquisition unit collects warehousing data based on sensors; the data transmission unit is used to transmit the collected data to a data processing center in real time; the data processing unit is used to build a big data processing platform; the digital twin model unit builds a high-precision digital twin model of warehousing facilities based on 3D modeling technology. The digital twin system and method for visualized warehousing based on big data analysis described in the present invention can reflect the operating status of the warehouse in real time and provide real-time decision support for managers. By establishing a three-dimensional model of the warehouse and simulating the operating process of the warehouse through simulation technology, it can provide managers with intuitive decision support and improve the operating efficiency of the warehouse.

Description

Technical Field

[0001] The present invention relates to the field of warehousing, and in particular to a digital twin system and method for visualized warehousing based on big data analysis. Background Art

[0002] In the modern logistics system, warehousing is one of the core parts of supply chain management. With the rapid development of information and intelligent technologies, enterprises' demand for warehousing management is no longer just traditional storage and distribution, but pays more attention to efficiency, accuracy and cost control. The integrated application of advanced technologies such as digital twins, 5G, Internet of Things, and artificial intelligence will provide strong technical support and solutions for the establishment of intelligent warehousing systems.

[0003] However, the existing warehousing system lacks a system that can display the warehouse operation status in real time, realize information interaction on real scenes, and realize the supervision of goods and equipment, and cannot optimize the decision-making of managers.

[0004] Therefore, it is necessary to propose a digital twin system and method for visual warehousing based on big data analysis to solve the above problems. Summary of the invention

[0005] The main purpose of the present invention is to provide a digital twin system and method for visual warehousing based on big data analysis, which can effectively solve the problems in the background technology.

[0006] To achieve the above object, the technical solution adopted by the present invention is:

[0007] A digital twin system for visual warehousing based on big data analysis includes a data acquisition unit, a data transmission unit, a data processing unit, a data analysis unit, and a digital twin model unit. The data acquisition unit collects warehouse data based on sensors; the data transmission unit is used to transmit the collected data to a data processing center in real time; the data processing unit cleans, organizes, stores and analyzes warehouse data based on a big data processing platform; the data analysis unit realizes intelligent warehousing scheduling based on artificial intelligence algorithms; the digital twin model unit constructs a high-precision digital twin model of warehousing facilities based on 3D modeling technology; the visual interaction unit develops a Web-based visual interaction interface, and users can access the digital twin system anytime and anywhere through terminal devices.

[0008] Preferably, the sensors include temperature and humidity sensors, inventory weight sensors, shelf pressure sensors, location tracking sensors, and surveillance cameras installed in the warehouse, which are used to comprehensively collect warehouse environment data, inventory status data, and cargo movement data.

[0009] Preferably, the data collection unit includes a data docking module for performing data docking with the enterprise resource planning and warehouse management system of the enterprise to obtain order information, goods attribute information, and personnel information.

[0010] Preferably, the data processing unit includes a big data processing platform and an optimization module;

[0011] The big data processing platform is used to clean, organize, store and analyze warehouse data;

[0012] The optimization module is used to optimize the inventory turnover rate and the cargo storage location, including an inventory turnover rate optimization module and a cargo storage location optimization module;

[0013] The inventory turnover optimization module includes:

[0014] The inventory collection and preprocessing module is used to collect a large amount of inventory-related data, including the purchase time, purchase quantity, sales time, sales quantity, and inventory cost of the goods, and clean the collected data to remove duplicate, erroneous, or incomplete data;

[0015] The first machine learning algorithm uses a module, but the inventory turnover rate has a linear relationship with time, price, and market demand. The linear regression algorithm is used to establish a linear regression model with time as the independent variable and inventory turnover rate as the dependent variable to predict the trend of inventory turnover rate over time; when the inventory turnover rate is affected by the complex influence of goods category, season, and promotional activities, the decision tree algorithm is used to divide the data into different subsets according to different factors, and each subset has its own inventory turnover rate characteristics; for more complex nonlinear relationships, neural networks are used. When the inventory turnover rate is affected by multiple interrelated factors, including but not limited to the combined influence of market trends, competitor activities, and economic environment, the neural network learns these complex patterns and predicts the inventory turnover rate based on this;

[0016] The model training and evaluation module divides the preprocessed data into a training set and a test set, with the training set accounting for 70% and the test set accounting for 30%. The training set is used to train the selected machine learning algorithm, and the parameters of the model are adjusted so that the model can accurately fit the training data. The trained model is evaluated using the test set.

[0017] Apply the model to optimize the decision-making module and formulate inventory management strategies based on the model's prediction results. When the model predicts that the inventory turnover rate of a certain type of goods will decrease, promotional activities will be taken in advance to reduce the purchase volume; when the inventory turnover rate is predicted to increase, the inventory will be increased;

[0018] The cargo storage location optimization module includes:

[0019] The data collection and feature engineering module collects the attribute data of goods, including but not limited to the size, weight, frequency of entry and exit, and shelf life of goods; warehouse layout information, including but not limited to the location, height, and capacity of shelves, performs feature engineering on the collected data, calculates the volume of goods, and standardizes the frequency of entry and exit to keep it within the reasonable value range set by itself;

[0020] The second machine learning algorithm selection module clusters the goods according to their attributes and uses the K-Means algorithm to classify the goods into different categories so that the goods of the same category have similar characteristics. The goods with high frequency of entering and leaving the warehouse are clustered into one category and stored near the warehouse entrance and exit to facilitate the rapid storage and retrieval of the goods. The warehouse storage location optimization is regarded as a decision-making process of an intelligent agent in the warehouse environment. The intelligent agent learns the optimal storage strategy by constantly trying to place the goods in different locations and receiving reward or penalty signals based on the storage and retrieval efficiency of the goods.

[0021] The location optimization module allocates goods to appropriate storage areas and uses the storage strategy obtained by reinforcement learning to guide the storage location arrangement of goods in real time. When new goods enter the warehouse, the goods are placed in the optimal location according to the strategy, and the storage location strategy is continuously adjusted as the warehouse operation changes.

[0022] Preferably, the digital twin model unit includes:

[0023] The physical structure building module creates an accurate model of the smart shelf based on high-precision 3D modeling software, including the shelf's frame structure, beams, and shelves. Each component is given real physical material properties, including but not limited to the density, strength, and elastic modulus of steel, so that its mechanical behavior can be accurately simulated in the digital twin model. Detailed dimensioning is added to the shelf model to ensure that the model is exactly the same size as the actual shelf, and the error is controlled within a very small range, including but not limited to the height, width, depth of the shelf, and the spacing between each layer, so as to ensure that key dimensional parameters can accurately reflect the actual situation;

[0024] The sensor and equipment modeling module integrates the virtual model of sensors on the shelf model. Temperature and humidity sensors, inventory weight sensors, shelf pressure sensors, location tracking sensors, and surveillance cameras are used to model the handling equipment on the shelf, including but not limited to automatic guided vehicle docking devices and conveyor connection components, to demonstrate their interactive interface with the shelf and their working range;

[0025] The cargo model building module builds a 3D model library of various types of cargo. Each cargo model has different shapes, sizes, weights, and packaging material attributes. The cargo model can automatically generate instances on the shelf based on inventory data and place them according to the actual storage location and stacking method. A virtual display of identification codes is added to the cargo model so that detailed information of the cargo can be obtained by scanning or identifying these identification codes in the digital twin model.

[0026] Functional and interactive design module: By connecting with the sensor data on the real shelf, the sensor virtual model in the digital twin model unit can display the collected data in real time. The weight sensor model intuitively displays the weight of the goods on each layer of the current shelf with dynamic numbers and color changes. When the weight exceeds the preset safety threshold, the model automatically sends a warning signal; the temperature and humidity sensor model updates the temperature and humidity values ​​of the area where the shelf is located in real time, and displays the historical change curve of temperature and humidity in the form of a chart. When the temperature and humidity exceed the appropriate range for cargo storage, the system automatically starts the virtual model of the associated control equipment to display its working status and the effect of adjusting the environmental parameters;

[0027] Cargo management and operation simulation module, managers can perform virtual warehousing, outbound and relocation operations on cargo based on digital twin model units;

[0028] Shelf status monitoring and early warning module, the digital twin model unit monitors the structural health status of the shelf in real time;

[0029] Intelligent analysis and optimization suggestion module, based on big data analysis and machine learning algorithms, the digital twin model unit conducts real-time analysis of shelf storage efficiency, cargo turnover rate, and equipment utilization rate. By analyzing the storage time and in-and-out frequency of different types of goods on the shelves, the model automatically generates optimization suggestions for cargo storage locations to improve overall warehousing operation efficiency. Based on historical order data and market demand forecasts, the model predicts shelf inventory capacity requirements in advance and provides replenishment suggestions and inventory adjustment strategies. When it is predicted that the demand for a certain type of goods will increase significantly, the model recommends increasing the inventory of the goods in advance and reserving appropriate shelf space for it.

[0030] The digital twin method for visual warehousing based on big data analysis includes the following steps:

[0031] S1: Install temperature and humidity sensors, inventory weight sensors, shelf pressure sensors, location tracking sensors, and surveillance cameras at appropriate locations in the warehouse to collect warehouse environment data, inventory status data, and cargo movement data;

[0032] S2: Optimize the inventory turnover rate of the warehouse based on the collected data. When the inventory turnover rate of a certain item is about to decrease, take promotional activities in advance and reduce the purchase volume; when the inventory turnover rate is predicted to increase, increase the inventory;

[0033] S3: Optimize the storage location of goods based on the collected data, store small-volume goods that are frequently in and out of the warehouse at the bottom of the shelf or near the entrance and exit; store large-volume goods that are infrequently in and out of the warehouse at high places or inside the warehouse;

[0034] S4: A warehouse digital twin model unit is established based on high-precision 3D modeling software and collected warehouse data. It also includes temperature and humidity sensors, inventory weight sensors, shelf pressure sensors, location tracking sensors, surveillance cameras, and handling equipment installed at the warehouse. The sensor virtual model in the digital twin model unit displays the collected data in real time and gives the digital twin model unit operability. When performing warehousing operations, the system automatically recommends the best storage location based on the properties of the goods and the storage rules of the shelves, and displays the transportation path and placement process of the goods from the warehouse entrance to the designated location in the model in an animated form. For outbound operations, When the goods to be shipped out are selected in the model, the model automatically updates the inventory data and the display status of the goods on the shelves, and determines whether the shelves are deformed, tilted, or overloaded by analyzing the weight sensors and displacement sensors. When an abnormality is found, the model highlights the fault area at the corresponding position and issues detailed early warning information, including the type of fault, possible causes, and recommended treatment measures. It monitors the operating status of the equipment on the shelves, including but not limited to the operating speed of the conveyor, the power and working status of the AGV. When the equipment fails or its performance deteriorates, the model promptly reminds the management personnel and provides the equipment maintenance history and links to relevant technical documents.

[0035] Compared with the prior art, the present invention provides a digital twin system and method for visual warehousing based on big data analysis, which has the following beneficial effects:

[0036] This digital twin system and method for visualized warehousing based on big data analysis can reflect the operating status of the warehouse in real time in a warehouse using the digital twin system, provide real-time decision support for managers, establish a three-dimensional model of the warehouse, and simulate the operation process of the warehouse through simulation technology to provide intuitive decision support for managers. It can realize intelligent control of various equipment and resources in the warehouse, provide decision support, and improve the operating efficiency and intelligence level of the warehouse. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 is a system block diagram of the present invention;

[0038] Figure 2is a system block diagram of a digital processing unit of the present invention;

[0039] Figure 3 It is a system block diagram of the digital twin model unit of the present invention. DETAILED DESCRIPTION

[0040] In order to make the technical means, creative features, objectives and effects achieved by the present invention easy to understand, the present invention is further explained below in conjunction with specific implementation methods.

[0041] Embodiment 1:

[0042] like Figure 1-Figure 3 As shown, the digital twin system for visual warehousing based on big data analysis includes a data acquisition unit, a data transmission unit, a data processing unit, a data analysis unit, and a digital twin model unit. The data acquisition unit collects warehouse data based on sensors; the data transmission unit is used to transmit the collected data to the data processing center in real time; the data processing unit cleans, organizes, stores and analyzes the warehouse data based on the construction of a big data processing platform; the data analysis unit realizes intelligent warehouse scheduling based on artificial intelligence algorithms; the digital twin model unit builds a high-precision digital twin model of warehouse facilities based on 3D modeling technology; the visual interaction unit develops a Web-based visual interaction interface, and users can access the digital twin system anytime and anywhere through terminal devices. Sensors include temperature and humidity sensors, inventory weight sensors, shelf pressure sensors, location tracking sensors, and surveillance cameras installed in the warehouse, which are used to comprehensively collect warehouse environment data, inventory status data, and cargo movement data. The data acquisition unit includes a data docking module, which is used to conduct data docking with the enterprise resource planning and warehouse management system of the enterprise to obtain order information, cargo attribute information, and personnel information;

[0043] The data processing unit includes a big data processing platform and an optimization module;

[0044] The big data processing platform is used to clean, organize, store and analyze warehouse data;

[0045] The optimization module is used to optimize the inventory turnover rate and the cargo storage location, including the inventory turnover rate optimization module and the cargo storage location optimization module;

[0046] The inventory turnover optimization module includes:

[0047] The inventory collection and preprocessing module is used to collect a large amount of inventory-related data, including the purchase time, purchase quantity, sales time, sales quantity, and inventory cost of the goods, and clean the collected data to remove duplicate, erroneous, or incomplete data;

[0048] The first machine learning algorithm uses a module, but the inventory turnover rate has a linear relationship with time, price, and market demand. The linear regression algorithm is used to establish a linear regression model with time as the independent variable and inventory turnover rate as the dependent variable to predict the trend of inventory turnover rate over time; when the inventory turnover rate is affected by the complex influence of goods category, season, and promotional activities, the decision tree algorithm is used to divide the data into different subsets according to different factors, and each subset has its own inventory turnover rate characteristics; for more complex nonlinear relationships, neural networks are used. When the inventory turnover rate is affected by multiple interrelated factors, including but not limited to the combined influence of market trends, competitor activities, and economic environment, the neural network learns these complex patterns and predicts the inventory turnover rate based on this;

[0049] The model training and evaluation module divides the preprocessed data into a training set and a test set, with the training set accounting for 70% and the test set accounting for 30%. The training set is used to train the selected machine learning algorithm, and the parameters of the model are adjusted so that the model can accurately fit the training data. The trained model is evaluated using the test set.

[0050] Apply the model to optimize the decision-making module and formulate inventory management strategies based on the model's prediction results. When the model predicts that the inventory turnover rate of a certain type of goods will decrease, promotional activities will be taken in advance to reduce the purchase volume; when the inventory turnover rate is predicted to increase, the inventory will be increased;

[0051] The cargo storage location optimization module includes:

[0052] The data collection and feature engineering module collects the attribute data of goods, including but not limited to the size, weight, frequency of entry and exit, and shelf life of goods; warehouse layout information, including but not limited to the location, height, and capacity of shelves, performs feature engineering on the collected data, calculates the volume of goods, and standardizes the frequency of entry and exit to keep it within the reasonable value range set by itself;

[0053] The second machine learning algorithm selection module clusters the goods according to their attributes and uses the K-Means algorithm to classify the goods into different categories so that the goods of the same category have similar characteristics. The goods with high frequency of entering and leaving the warehouse are clustered into one category and stored near the warehouse entrance and exit to facilitate the rapid storage and retrieval of the goods. The warehouse storage location optimization is regarded as a decision-making process of an intelligent agent in the warehouse environment. The intelligent agent learns the optimal storage strategy by constantly trying to place the goods in different locations and receiving reward or penalty signals based on the storage and retrieval efficiency of the goods.

[0054] The location optimization module allocates goods to appropriate storage areas and uses the storage strategy obtained by reinforcement learning to guide the storage location arrangement of goods in real time. When new goods enter the warehouse, the goods are placed in the optimal location according to the strategy, and the storage location strategy is continuously adjusted as the warehouse operation changes.

[0055] The digital twin model unit includes:

[0056] The physical structure building module creates an accurate model of the smart shelf based on high-precision 3D modeling software, including the shelf's frame structure, beams, and shelves. Each component is given real physical material properties, including but not limited to the density, strength, and elastic modulus of steel, so that its mechanical behavior can be accurately simulated in the digital twin model. Detailed dimensioning is added to the shelf model to ensure that the model is exactly the same size as the actual shelf, and the error is controlled within a very small range, including but not limited to the height, width, depth of the shelf, and the spacing between each layer, so as to ensure that key dimensional parameters can accurately reflect the actual situation;

[0057] The sensor and equipment modeling module integrates the virtual model of sensors on the shelf model. Temperature and humidity sensors, inventory weight sensors, shelf pressure sensors, location tracking sensors, and surveillance cameras are used to model the handling equipment on the shelf, including but not limited to automatic guided vehicle docking devices and conveyor connection components, to demonstrate their interactive interface with the shelf and their working range;

[0058] The cargo model building module builds a 3D model library of various types of cargo. Each cargo model has different shapes, sizes, weights, and packaging material attributes. The cargo model can automatically generate instances on the shelf based on inventory data and place them according to the actual storage location and stacking method. A virtual display of identification codes is added to the cargo model so that detailed information of the cargo can be obtained by scanning or identifying these identification codes in the digital twin model.

[0059] Functional and interactive design module: By connecting with the sensor data on the real shelf, the sensor virtual model in the digital twin model unit can display the collected data in real time. The weight sensor model intuitively displays the weight of the goods on each layer of the current shelf with dynamic numbers and color changes. When the weight exceeds the preset safety threshold, the model automatically sends a warning signal; the temperature and humidity sensor model updates the temperature and humidity values ​​of the area where the shelf is located in real time, and displays the historical change curve of temperature and humidity in the form of a chart. When the temperature and humidity exceed the appropriate range for cargo storage, the system automatically starts the virtual model of the associated control equipment to display its working status and the effect of adjusting the environmental parameters;

[0060] Cargo management and operation simulation module, managers can perform virtual warehousing, outbound and relocation operations on cargo based on digital twin model units;

[0061] Shelf status monitoring and early warning module, the digital twin model unit monitors the structural health status of the shelf in real time;

[0062] Intelligent analysis and optimization suggestion module, based on big data analysis and machine learning algorithms, the digital twin model unit conducts real-time analysis of shelf storage efficiency, cargo turnover rate, and equipment utilization rate. By analyzing the storage time and in-and-out frequency of different types of goods on the shelves, the model automatically generates optimization suggestions for cargo storage locations to improve overall warehousing operation efficiency. Based on historical order data and market demand forecasts, the model predicts shelf inventory capacity requirements in advance and provides replenishment suggestions and inventory adjustment strategies. When it is predicted that the demand for a certain type of goods will increase significantly, the model recommends increasing the inventory of the goods in advance and reserving appropriate shelf space for it.

[0063] Embodiment 2:

[0064] The digital twin method for visual warehousing based on big data analysis includes the following steps:

[0065] S1: Install temperature and humidity sensors, inventory weight sensors, shelf pressure sensors, location tracking sensors, and surveillance cameras at appropriate locations in the warehouse to collect warehouse environment data, inventory status data, and cargo movement data;

[0066] S2: Optimize the inventory turnover rate of the warehouse based on the collected data. When the inventory turnover rate of a certain item is about to decrease, take promotional activities in advance and reduce the purchase volume; when the inventory turnover rate is predicted to increase, increase the inventory;

[0067] S3: Optimize the storage location of goods based on the collected data, store small-volume goods that are frequently in and out of the warehouse at the bottom of the shelf or near the entrance and exit; store large-volume goods that are infrequently in and out of the warehouse at high places or inside the warehouse;

[0068] S4: A warehouse digital twin model unit is established based on high-precision 3D modeling software and collected warehouse data. It also includes temperature and humidity sensors, inventory weight sensors, shelf pressure sensors, location tracking sensors, surveillance cameras, and handling equipment installed at the warehouse. The sensor virtual model in the digital twin model unit displays the collected data in real time and gives the digital twin model unit operability. When performing warehousing operations, the system automatically recommends the best storage location based on the properties of the goods and the storage rules of the shelves, and displays the transportation path and placement process of the goods from the warehouse entrance to the designated location in the model in an animated form. For outbound operations, When the goods to be shipped out are selected in the model, the model automatically updates the inventory data and the display status of the goods on the shelves, and determines whether the shelves are deformed, tilted, or overloaded by analyzing the weight sensors and displacement sensors. When an abnormality is found, the model highlights the fault area at the corresponding position and issues detailed early warning information, including the type of fault, possible causes, and recommended treatment measures. It monitors the operating status of the equipment on the shelves, including but not limited to the operating speed of the conveyor, the power and working status of the AGV. When the equipment fails or its performance deteriorates, the model promptly reminds the management personnel and provides the equipment maintenance history and links to relevant technical documents.

[0069] The above shows and describes the basic principles and main features of the present invention and the advantages of the present invention. It should be understood by those skilled in the art that the present invention is not limited to the above embodiments. The above embodiments and descriptions are only for explaining the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, which fall within the scope of the present invention to be protected. The scope of protection of the present invention is defined by the attached claims and their equivalents.

Claims

1. A digital twin system for visual warehousing based on big data analysis, including a data acquisition unit, a data transmission unit, a data processing unit, a data analysis unit, and a digital twin model unit, characterized in that: The data acquisition unit collects warehouse data based on sensors; the data transmission unit is used to transmit the collected data to the data processing center in real time; the data processing unit cleans, organizes, stores and analyzes warehouse data based on the construction of a big data processing platform; the data analysis unit realizes intelligent warehouse scheduling based on artificial intelligence algorithms; the digital twin model unit constructs a high-precision digital twin model of storage facilities based on 3D modeling technology; the visualization interaction unit develops a Web-based visualization interaction interface, and users can access the digital twin system anytime and anywhere through terminal devices.

2. The digital twin system for visual warehousing based on big data analysis according to claim 1 is characterized by: The sensors include temperature and humidity sensors, inventory weight sensors, shelf pressure sensors, location tracking sensors, and surveillance cameras installed in the warehouse, which are used to comprehensively collect warehouse environment data, inventory status data, and cargo movement data.

3. The digital twin system for visual warehousing based on big data analysis according to claim 1 is characterized by: The data collection unit includes a data docking module, which is used to perform data docking with the enterprise resource planning and warehouse management system of the enterprise to obtain order information, goods attribute information, and personnel information.

4. The digital twin system for visual warehousing based on big data analysis according to claim 1 is characterized by: The data processing unit includes a big data processing platform and an optimization module; The big data processing platform is used to clean, organize, store and analyze warehouse data; The optimization module is used to optimize the inventory turnover rate and the cargo storage location, including an inventory turnover rate optimization module and a cargo storage location optimization module; The inventory turnover optimization module includes: The inventory collection and preprocessing module is used to collect a large amount of inventory-related data, including the purchase time, purchase quantity, sales time, sales quantity, and inventory cost of the goods, and clean the collected data to remove duplicate, erroneous, or incomplete data; The first machine learning algorithm uses a module, but the inventory turnover rate has a linear relationship with time, price, and market demand. The linear regression algorithm is used to establish a linear regression model with time as the independent variable and inventory turnover rate as the dependent variable to predict the trend of inventory turnover rate over time; when the inventory turnover rate is affected by the complex influence of goods category, season, and promotional activities, the decision tree algorithm is used to divide the data into different subsets according to different factors, and each subset has its own inventory turnover rate characteristics; for more complex nonlinear relationships, neural networks are used. When the inventory turnover rate is affected by multiple interrelated factors, including but not limited to the combined influence of market trends, competitor activities, and economic environment, the neural network learns these complex patterns and predicts the inventory turnover rate based on this; The model training and evaluation module divides the preprocessed data into a training set and a test set, with the training set accounting for 70% and the test set accounting for 30%. The training set is used to train the selected machine learning algorithm, and the parameters of the model are adjusted so that the model can accurately fit the training data. The trained model is evaluated using the test set. Apply the model to optimize the decision-making module and formulate inventory management strategies based on the model's prediction results. When the model predicts that the inventory turnover rate of a certain type of goods will decrease, promotional activities will be taken in advance to reduce the purchase volume; when the inventory turnover rate is predicted to increase, the inventory will be increased; The cargo storage location optimization module includes: The data collection and feature engineering module collects the attribute data of goods, including but not limited to the size, weight, frequency of entry and exit, and shelf life of goods; warehouse layout information, including but not limited to the location, height, and capacity of shelves, performs feature engineering on the collected data, calculates the volume of goods, and standardizes the frequency of entry and exit to keep it within the reasonable value range set by itself; The second machine learning algorithm selection module clusters the goods according to their attributes and uses the K-Means algorithm to classify the goods into different categories so that the goods of the same category have similar characteristics. The goods with high frequency of entering and leaving the warehouse are clustered into one category and stored near the warehouse entrance and exit to facilitate the rapid storage and retrieval of the goods. The warehouse storage location optimization is regarded as a decision-making process of an intelligent agent in the warehouse environment. The intelligent agent learns the optimal storage strategy by constantly trying to place the goods in different locations and receiving reward or penalty signals based on the storage and retrieval efficiency of the goods. The location optimization module allocates goods to appropriate storage areas and uses the storage strategy obtained by reinforcement learning to guide the storage location arrangement of goods in real time. When new goods enter the warehouse, the goods are placed in the optimal location according to the strategy, and the storage location strategy is continuously adjusted as the warehouse operation changes.

5. The digital twin system for visual warehousing based on big data analysis according to claim 1 is characterized by: The digital twin model unit includes: The physical structure building module creates an accurate model of the smart shelf based on high-precision 3D modeling software, including the shelf's frame structure, beams, and shelves. Each component is given real physical material properties, including but not limited to the density, strength, and elastic modulus of steel, so that its mechanical behavior can be accurately simulated in the digital twin model. Detailed dimensioning is added to the shelf model to ensure that the model is exactly the same size as the actual shelf, and the error is controlled within a very small range, including but not limited to the height, width, depth of the shelf, and the spacing between each layer, so as to ensure that key dimensional parameters can accurately reflect the actual situation; The sensor and equipment modeling module integrates the virtual model of sensors on the shelf model. Temperature and humidity sensors, inventory weight sensors, shelf pressure sensors, location tracking sensors, and surveillance cameras are used to model the handling equipment on the shelf, including but not limited to automatic guided vehicle docking devices and conveyor connection components, to demonstrate their interactive interface with the shelf and their working range; The cargo model building module builds a 3D model library of various types of cargo. Each cargo model has different shapes, sizes, weights, and packaging material attributes. The cargo model can automatically generate instances on the shelf based on inventory data and place them according to the actual storage location and stacking method. A virtual display of identification codes is added to the cargo model so that detailed information of the cargo can be obtained by scanning or identifying these identification codes in the digital twin model. Functional and interactive design module: By connecting with the sensor data on the real shelf, the sensor virtual model in the digital twin model unit can display the collected data in real time. The weight sensor model intuitively displays the weight of the goods on each layer of the current shelf with dynamic numbers and color changes. When the weight exceeds the preset safety threshold, the model automatically sends a warning signal; the temperature and humidity sensor model updates the temperature and humidity values ​​of the area where the shelf is located in real time, and displays the historical change curve of temperature and humidity in the form of a chart. When the temperature and humidity exceed the appropriate range for cargo storage, the system automatically starts the virtual model of the associated control equipment to display its working status and the effect of adjusting the environmental parameters; Cargo management and operation simulation module, managers can perform virtual warehousing, outbound and relocation operations on cargo based on digital twin model units; Shelf status monitoring and early warning module, the digital twin model unit monitors the structural health status of the shelf in real time; Intelligent analysis and optimization suggestion module, based on big data analysis and machine learning algorithms, the digital twin model unit conducts real-time analysis of shelf storage efficiency, cargo turnover rate, and equipment utilization rate. By analyzing the storage time and in-and-out frequency of different types of goods on the shelves, the model automatically generates optimization suggestions for cargo storage locations to improve overall warehousing operation efficiency. Based on historical order data and market demand forecasts, the model predicts shelf inventory capacity requirements in advance and provides replenishment suggestions and inventory adjustment strategies. When it is predicted that the demand for a certain type of goods will increase significantly, the model recommends increasing the inventory of the goods in advance and reserving appropriate shelf space for it.

6. A digital twin method for visualized warehousing based on big data analysis, using the digital twin system for visualized warehousing based on big data analysis as described in any of claims 1 to 5, characterized in that: The steps include: S1: Install temperature and humidity sensors, inventory weight sensors, shelf pressure sensors, location tracking sensors, and surveillance cameras at appropriate locations in the warehouse to collect warehouse environment data, inventory status data, and cargo movement data; S2: Optimize the inventory turnover rate of the warehouse based on the collected data. When the inventory turnover rate of a certain item is about to decrease, take promotional activities in advance and reduce the purchase volume; When inventory turnover is predicted to increase, increase inventory; S3: Optimize the storage location of goods based on the collected data, store small-volume goods that are frequently in and out of the warehouse at the bottom of the shelf or near the entrance and exit; store large-volume goods that are infrequently in and out of the warehouse at high places or inside the warehouse; S4: A warehouse digital twin model unit is established based on high-precision 3D modeling software and collected warehouse data. It also includes temperature and humidity sensors, inventory weight sensors, shelf pressure sensors, location tracking sensors, surveillance cameras, and handling equipment installed at the warehouse. The sensor virtual model in the digital twin model unit displays the collected data in real time and gives the digital twin model unit operability. When performing warehousing operations, the system automatically recommends the best storage location based on the properties of the goods and the storage rules of the shelves, and displays the transportation path and placement process of the goods from the warehouse entrance to the designated location in the model in an animated form. For outbound operations, When the goods to be shipped out are selected in the model, the model automatically updates the inventory data and the display status of the goods on the shelves, and determines whether the shelves are deformed, tilted, or overloaded by analyzing the weight sensors and displacement sensors. When an abnormality is found, the model highlights the fault area at the corresponding position and issues detailed early warning information, including the type of fault, possible causes, and recommended treatment measures. It monitors the operating status of the equipment on the shelves, including but not limited to the operating speed of the conveyor, the power and working status of the AGV. When the equipment fails or its performance deteriorates, the model promptly reminds the management personnel and provides the equipment maintenance history and links to relevant technical documents.

Images