Warehouse logistics visual monitoring system based on digital twinning

Abstract

The invention discloses a warehouse logistics visual monitoring system based on digital twinning, and belongs to the technical field of warehouse logistics fault monitoring, and the system specifically comprises the steps: collecting various data in a warehouse environment, including a warehouse architecture, an article position, inventory information, equipment state data and personnel activities, carrying out the preprocessing, and obtaining a database; constructing a hierarchical digital twinborn model and a constructed virtual-real space real-time mapping system on the basis of various data in the preprocessed storage environment, performing simulation prediction on operation of storage equipment by using the digital twinborn model, performing multi-dimensional monitoring on storage logistics, constructing a fault prediction model, and predicting and analyzing faults of the storage equipment; according to the method, virtual-real compensation optimization is carried out by using Kalman filtering and an optimization algorithm, and the precision of the digital twin model is improved, so that potential fault risks can be identified in advance, early warning can be given out in time, the equipment fault rate is greatly reduced, the maintenance cost is reduced, and the normal operation time of equipment is prolonged.

Description

Technical Field

[0001] The present invention belongs to the technical field of warehousing and logistics fault monitoring, and specifically to a warehousing and logistics visualization monitoring system based on digital twins. Background Art

[0002] With the rapid development of the logistics industry, modern warehouse management has become increasingly complex. Efficient and accurate management of goods and equipment has become a pressing issue, especially in large-scale, intelligent warehouses. Traditional warehouse logistics monitoring methods, which rely heavily on manual management and empirical judgment, have limitations. These methods typically only enable real-time monitoring of a single piece of equipment or a small amount of data within the warehouse, but are unable to comprehensively and accurately monitor and optimize the entire warehouse logistics system.

[0003] Digital twins, a technology based on the mapping of the physical world to virtual models, can be used in the field of warehousing and logistics to map a warehouse's physical space, equipment operation, cargo status, environmental conditions, and other elements into the virtual world in real time, creating a dynamic, comprehensive, and real-time virtual model. This virtual model not only reflects the real-time state of the actual warehouse environment but also enables simulation and prediction, further optimizing warehouse operations. However, there are some drawbacks: ensuring the accuracy of digital twin models and the precision of virtual-physical data transmission across a variety of equipment and warehousing and logistics scenarios remains a technical challenge.

[0004] For example, the patent application with publication number CN119005858A discloses a material warehouse management system based on digital twin technology, which includes a data acquisition and monitoring module, a digital twin warehouse model module, an inventory management module, a logistics and transportation management module, an order and demand forecasting module, a warehouse operation management module, a safety and risk management module, and a data analysis and decision support module. The data acquisition and monitoring module collects and monitors warehouse environment and material status data in real time, the digital twin warehouse model module dynamically simulates the actual warehouse situation, the inventory management module manages inventory recording, tracking, classification, and automatic replenishment, the logistics and transportation management module manages the material in and out process, the order and demand forecasting module processes orders and conducts demand analysis, the warehouse operation management module manages personnel and equipment, the safety and risk management module monitors warehouse safety, and the data analysis and decision support module provides optimized decision support. This technical solution improves the efficiency, safety, and accuracy of material warehousing.

[0005] For example, the patent application with publication number CN118229182A discloses a warehousing and logistics workshop monitoring system and method based on digital twin technology, including a user login module, a 3D interactive module, an early warning module, a roaming module, a video monitoring module, a digital workshop module, an equipment management module, a production efficiency module, an inventory management module and a data analysis dashboard module; with the help of digital twin technology, a digital twin of the warehousing and logistics workshop is constructed, and the real-time synchronization and one-to-one mapping of equipment status and operation status are realized in the digital twin. With the help of data collection and data analysis, the visualization of workshop operation data and production transparency are realized, cross-system data fusion and aggregation are realized, the interconnection of various business links is realized, the expression of all elements of the workshop and global visualization are realized, so as to better monitor the real-time status of the workshop and improve production efficiency.

[0006] The defect of the above technical solution is that it does not take into account the possibility of certain errors in data transmission between virtual and real spaces. Summary of the Invention

[0007] In response to the shortcomings of the existing technology, the present invention proposes a warehouse logistics visualization monitoring system based on digital twins. By combining advanced technologies such as digital twins, the Internet of Things, data analysis and artificial intelligence, virtual-real compensation is performed on data transmission between virtual and real spaces, realizing all-round, real-time and dynamic visualization monitoring of the warehouse logistics environment, and improving the accuracy of warehouse equipment fault prediction.

[0008] To achieve the above object, the present invention provides the following technical solutions:

[0009] The digital twin-based warehouse logistics visualization monitoring system includes: data acquisition module, digital twin construction module, virtual-reality mapping module and prediction analysis module;

[0010] The data collection module is used to collect various data in the warehouse environment, including: warehouse structure, item location, inventory information, equipment status data and personnel activities, and perform pre-processing;

[0011] The digital twin construction module constructs a hierarchical digital twin model based on the pre-processed data in the warehouse environment;

[0012] The virtual-reality mapping module is used to build a real-time mapping system between virtual and real space, and use the digital twin model to simulate and predict the operation of storage equipment;

[0013] The prediction and analysis module is used to monitor warehousing logistics in multiple dimensions, build a fault prediction model, and predict and analyze the failure of warehousing equipment.

[0014] Specifically, the digital twin construction module includes: a first-layer digital twin model construction unit, a second-layer digital twin model construction unit, and a third-layer digital twin model construction unit;

[0015] The first-layer digital twin model construction unit is used to build a device-level digital twin, model the parameters of the device under different operating conditions, make adjustments based on data, and provide real-time feedback on the operating status of the device;

[0016] The second-layer digital twin model construction unit is used to build a digital twin at the cargo location level, assigning different color heat values to different cargo locations based on the cargo's residence time and turnover frequency indicators in the cargo location;

[0017] The third-layer digital twin model construction unit is used to build a system-level digital twin to optimize the overall logistics path planning within the warehouse.

[0018] Specifically, the virtual-reality mapping module includes: a virtual-reality mapping unit and a state prediction unit;

[0019] The virtual-reality mapping unit is used to synchronize data between the storage equipment and the digital twin model through the Internet of Things technology and to compensate for virtual-reality deviations;

[0020] The state prediction unit is used to predict the operating state of the storage equipment.

[0021] Specifically, the virtual-real deviation compensation includes:

[0022] The Kalman filter is used to dynamically filter the virtual and real space data to obtain preliminary correction data for the deviation correction of the virtual and real space data;

[0023] The preliminary correction data is used as the initial population, and the fitness function is defined as the weighted sum of the storage equipment state deviation and the path planning efficiency. Crossover individuals are selected based on the fitness, and the elite individuals are dynamically adjusted. When the storage equipment is in the startup phase and the state volatility is greater than 15%, the top n% of individuals are retained as crossover individuals. When the storage equipment is in the stable phase and the state volatility is less than 5%, m% of crossover individuals are retained, satisfying m>n.

[0024] Adaptive crossover iteration is used. When the storage equipment is in the startup phase and the state volatility is greater than 15%, the crossover probability is adjusted to 0.8. When the storage equipment is in the stable phase and the state volatility is less than 5%, the crossover probability is adjusted to 0.3. If the optimal solution is not updated after 20 iterations or the fitness variance is less than the set threshold for 50 consecutive iterations, part of the population is reset.

[0025] Randomly select 30% of non-crossover individuals, perform completely random initialization on the selected non-crossover individuals or generate new individuals based on historical data, and perform crossover mutation operations on the randomly initialized individuals or the generated new individuals;

[0026] When convergence or the maximum evolutionary generation is reached, the optimal deviation correction solution is output.

[0027] Specifically, the virtual-real space data is the data of the interaction between warehousing and goods and the digital twin model.

[0028] Specifically, the predicting the operating status of the storage equipment includes:

[0029] Based on the pre-processed equipment status data and the first-layer digital twin model, the real-time state equation of the storage equipment is established;

[0030] Update the real-time status of the storage equipment according to the status change rate of the storage equipment within △t;

[0031] The digital twin model is used on the cloud server to simulate the operating status of the storage equipment in real time, and the future status of the storage equipment is predicted by inputting the current status and input signals.

[0032] Specifically, the prediction and analysis module includes: a fault prediction model construction and training unit and a prediction and analysis unit;

[0033] The fault prediction model construction training unit is used to construct a fault prediction model and perform training;

[0034] The prediction and analysis unit uses the trained fault prediction model to predict and analyze the faults of the storage equipment.

[0035] Specifically, the method of using the trained fault prediction model to predict and analyze storage equipment failures includes:

[0036] Select a machine learning algorithm to build a fault prediction model, using a labeled dataset D train Train the fault prediction model and set the model training objective to minimize the loss function;

[0037] The fault prediction model is trained until the minimization loss function of the model's training objective converges and remains unchanged, and the training is stopped to obtain a trained fault prediction model.

[0038] Specifically, the trained fault prediction model is used to predict and analyze storage equipment failures, including:

[0039] Extract the key features from the historical data set D of storage equipment. The key features include statistical features and frequency domain features, and select the feature set F that has the best discriminative power for fault judgment. new, F new ={Ft1,Ft2,...,Ft m}, Ft m It represents the mth feature with the best discriminative ability for fault judgment;

[0040] The extracted feature set F new Input the trained storage equipment fault identification model to obtain the predicted probability of storage equipment failure;

[0041] Set the fault warning threshold Ψ according to actual needs. If the predicted probability of storage equipment failure is greater than Ψ, an alarm is triggered; otherwise, no alarm is triggered.

[0042] Compared with the prior art, the present invention has the following beneficial effects:

[0043] 1. The present invention proposes a digital twin-based warehouse logistics visualization monitoring system. By constructing a digital twin model of the warehouse logistics system and using Kalman filtering and optimization algorithms to perform virtual-reality compensation optimization, it can monitor all aspects of the warehouse in real time, including cargo storage, equipment operation, logistics routes and other information, thereby improving the accuracy of the digital twin model.

[0044] 2. This invention proposes a digital twin-based warehouse logistics visualization monitoring system. Based on real-time collected data, combined with digital twin models and artificial intelligence algorithms, it can intelligently optimize various operations in the warehouse logistics system, including: equipment scheduling and route planning, real-time tracking of inventory status, and dynamic adjustment of cargo storage locations, thereby reducing redundancy in transportation routes, optimizing resource allocation, and improving logistics efficiency.

[0045] 3. The present invention proposes a warehouse logistics visualization monitoring system based on digital twins. Through continuous monitoring and analysis of equipment operation data, the digital twin model can realize real-time evaluation and prediction of equipment status. Combined with big data analysis and machine learning algorithms, it can identify potential failure risks in advance and issue early warnings in time to guide warehouse managers to perform maintenance and replacement operations, significantly reducing equipment failure rates, reducing maintenance costs, and improving equipment uptime. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 This is the architecture diagram of the digital twin-based warehouse logistics visualization monitoring system provided by the present invention;

[0047] Figure 2 The present invention provides a flow chart for correcting deviations in virtual and real space data. DETAILED DESCRIPTION

[0048] The present application is described in detail below with reference to specific embodiments. The following embodiments will help those skilled in the art to further understand the present application, but are not intended to limit the present application in any form. It should be noted that those skilled in the art may make several variations and improvements without departing from the scope of the present application. These all fall within the scope of protection of the present application.

[0049] In order to make the purpose, technical solutions and advantages of this application more clear, the following further describes this application in detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain this application and are not intended to limit this application.

[0050] It should be noted that, if there is no conflict, the various features in the embodiments of the present application can be combined with each other and are all within the scope of protection of the present application. In addition, although the functional modules are divided in the device schematic and the logical order is shown in the flow chart, in some cases, the steps shown or described can be performed in a different order than the module division in the device or the order in the flow chart. In addition, the words "first", "second", "third", etc. used in this application do not limit the data and execution order, but only distinguish between the same items or similar items with basically the same functions and effects.

[0051] Unless otherwise defined, all technical and scientific terms used in this specification have the same meanings as those commonly understood by those skilled in the art to which this application belongs. The terms used in this specification and in the specification of this application are only for the purpose of describing specific embodiments and are not intended to limit this application. The term "and / or" as used in this specification includes any and all combinations of one or more of the relevant listed items.

[0052] Example 1

[0053] See also Figure 1 、 Figure 2 , an embodiment provided by the present invention: a warehouse logistics visualization monitoring system based on digital twins, comprising: a data acquisition module, a digital twin construction module, a virtual-reality mapping module and a prediction analysis module;

[0054] The data collection module is used to collect various data in the warehouse environment, including: warehouse structure, item location, inventory information, equipment status data and personnel activities, and perform pre-processing;

[0055] Use high-precision mapping drones, lidar, sensors, RFID, cameras and other equipment to collect various data in the warehouse environment;

[0056] The preprocessing includes: denoising, cleaning the data, removing abnormal data and redundant data, denoising the image, and removing noise points in the image;

[0057] The digital twin construction module constructs a hierarchical digital twin model based on the pre-processed data in the warehouse environment;

[0058] The digital twin construction module includes: a first-layer digital twin model construction unit, a second-layer digital twin model construction unit, and a third-layer digital twin model construction unit;

[0059] The first-layer digital twin model construction unit is used to build a device-level digital twin, model the parameters of the device under different operating conditions, make adjustments based on data, and provide real-time feedback on the operating status of the device;

[0060] The device-level digital twin is constructed to model the parameters of the device under different operating conditions, make adjustments based on data, and provide real-time feedback on the device's operating status, including:

[0061] Deploy sensors such as accelerometers, displacement sensors, and force sensors on equipment such as AGVs and stackers to collect real-time data on the equipment's motion status, speed, load information, working environment data (e.g., temperature, vibration, speed), and status data (e.g., running, standby, fault, etc.);

[0062] Based on the mechanical characteristics and kinematic principles of the equipment, dynamic models of equipment such as AGVs and stackers are constructed. Parameters of the equipment under different operating conditions (such as current, speed, acceleration, etc.) are modeled to ensure that the equipment twin can reflect the actual operating status of the equipment in real time.

[0063] For example: using Newtonian mechanics to establish the dynamic equations of AGV, simulating the motion trajectory of the equipment, load mechanical characteristics, etc.

[0064] Utilizing historical data and sensor data, machine learning algorithms (such as regression analysis and deep neural networks) are used to model the movement process of the equipment. Based on data-driven model adjustments, real-time feedback on the operating status of the equipment is provided, and the differences between the physical model and the actual situation are automatically corrected.

[0065] The second-layer digital twin model construction unit is used to build a digital twin at the cargo location level, assigning different color heat values to different cargo locations based on indicators such as the cargo's residence time and turnover frequency in the cargo location;

[0066] The digital twin at the cargo location level is constructed, and different locations are assigned different heat values based on indicators such as the residence time and turnover frequency of the goods in the cargo location, including:

[0067] Through technologies such as RFID and barcode scanning, the location information and storage status (e.g., taken, waiting for replenishment, etc.) of each item in the warehouse are obtained in real time. Environmental sensors (such as temperature and humidity sensors) are used to collect storage environment data to determine the impact of storage conditions on the logistics process.

[0068] Using cargo location information and environmental data, combined with area division and cargo storage density, and using a heat map generation algorithm, real-time heat maps of different areas within the warehouse are generated;

[0069] In this embodiment, the heat map reflects the storage density and activity status of goods in different areas of the warehouse. For example, a certain area may cause a logistics bottleneck due to excessive storage of goods, or an abnormal storage environment due to factors such as high temperature or humidity;

[0070] Update heat maps based on real-time data and dynamically adjust cargo storage locations to avoid inefficiencies or congestion in logistics operations caused by improper spatial layout.

[0071] The third-layer digital twin model construction unit is used to build a system-level digital twin to optimize the overall logistics path planning within the warehouse.

[0072] The system-level digital twin is constructed to optimize the overall logistics path planning within the warehouse, including:

[0073] Based on the first-layer digital twin model and the second-layer digital twin model, a global model of warehousing logistics is constructed to reflect the interrelationships and dynamic changes among various warehousing equipment, goods, and warehouse areas;

[0074] Use path planning algorithms (such as A* algorithm, Dijkstra algorithm, etc.) to optimize the logistics paths of warehousing equipment such as AGVs and stackers, and automatically select the optimal path by dynamically calculating the shortest distance, minimum waiting time and other indicators of the logistics path;

[0075] In this embodiment, taking into account real-time changes in the warehouse (e.g., changes in the storage location of goods, changes in the operating status of equipment), the optimization algorithm can update the logistics path in real time under different circumstances to avoid congestion or inefficient transportation routes;

[0076] A multi-objective scheduling algorithm is used to balance multiple objectives such as equipment load, path length, and transportation time. The working order of AGVs, stackers, and other equipment is intelligently scheduled based on the equipment's working status, priority, and real-time needs.

[0077] In this embodiment, the construction of a hierarchical twin model architecture can provide real-time, dynamic monitoring and optimization support in warehousing and logistics management, and model and optimize the warehousing and logistics environment from the equipment layer, cargo storage layer and global logistics system layer respectively, ensuring the efficiency, real-time and intelligence of the digital twin in the warehousing and logistics process. It can not only improve the operating efficiency of warehousing and logistics, but also reduce manual intervention, predict and avoid potential problems in advance.

[0078] The virtual-reality mapping module is used to build a real-time mapping system between virtual and real space, and use the digital twin model to simulate and predict the operation of storage equipment;

[0079] The virtual-real mapping module includes: a virtual-real mapping unit and a state prediction unit;

[0080] The virtual-reality mapping unit synchronizes data between the storage equipment and the digital twin model through the Internet of Things technology and compensates for virtual-reality deviations;

[0081] The virtual-real deviation compensation includes:

[0082] Using Kalman filtering to dynamically filter the virtual and real space data to obtain preliminary correction data for the deviation correction of the virtual and real space data. The virtual and real space data is the data of the interaction between the warehouse and the goods and the digital twin model;

[0083] The preliminary correction data is used as the initial population, and the fitness function is defined as the weighted sum of the storage equipment state deviation and the path planning efficiency. Crossover individuals are selected based on the fitness, and the elite individuals are dynamically adjusted. When the storage equipment is in the startup phase and the state volatility is greater than 15%, the top n% of individuals are retained as crossover individuals. When the storage equipment is in the stable phase and the state volatility is less than 5%, m% of crossover individuals are retained, satisfying m>n.

[0084] In this embodiment, when the storage equipment is in the startup phase and the state fluctuation rate is greater than 15%, n may be between 5 and 10; when the storage equipment is in the stable phase and the state fluctuation rate is less than 5%, m may be between 15 and 20;

[0085] Adaptive crossover iteration is used. When the storage equipment is in the startup phase and the state volatility is greater than 15%, the crossover probability is adjusted to 0.8. When the storage equipment is in the stable phase and the state volatility is less than 5%, the crossover probability is adjusted to 0.3. If the optimal solution is not updated after 20 iterations or the fitness variance is less than the set threshold for 50 consecutive iterations, part of the population is reset.

[0086] Randomly select 30% of non-crossover individuals, perform completely random initialization on the selected non-crossover individuals or generate new individuals based on historical data, and perform crossover mutation operations on the randomly initialized individuals or the generated new individuals;

[0087] In this embodiment, when an equipment anomaly is detected (e.g., an AGV frequently deviates from the planned path), part of the population is reset, 30% of the AGV path planning individuals are reset, a new path is generated based on real-time sensor data, and iteration is performed to select the optimal solution. The 30% here is not a fixed value and can be 25%-30%;

[0088] When convergence or the maximum evolutionary generation is reached, the optimal deviation correction solution is output.

[0089] In this embodiment, during the real-time data mapping process between virtual and physical spaces, deviations may occur between the virtual twin and the actual device due to sensor errors, device delays, network fluctuations, and other factors. These sources of deviation are identified and analyzed, focusing on aspects such as the device motion trajectory, cargo location, and sensor reading errors in the physical environment. Kalman filtering optimizes the state estimation of devices and cargo in the virtual model by comparing real-time sensor data with the expected state of the digital twin, correcting the virtual-real deviation in real time. However, when data fluctuates significantly, its accuracy is significantly insufficient.

[0090] The genetic algorithm has high requirements for the selection of the initial population. The quality of the initial population has a great impact on the results. Kalman filtering is used to solve the problem of the quality of the initial population. In the crossover iteration process, it is easy to fall into local optimality or converge prematurely. This application adopts adaptive crossover iteration. In the early stage of iteration, when the individual differences in the population are large, the crossover probability is appropriately increased to promote the widespread dissemination of excellent genes and quickly explore the solution space. As the iteration deepens, the population tends to be homogenized, the crossover probability is reduced, and local potential better solutions are finely mined. For example, taking the compensation of storage equipment operating parameter deviation as an example, the parameters fluctuate greatly during the equipment startup phase. A high crossover probability (0.8) helps to quickly locate the appropriate parameter range. After the operation is stable, fine-tuning is performed and a low crossover probability (0.3) is used to ensure accurate compensation of the deviation. If the crossover probability is a fixed value, it is easy to converge prematurely. This application achieves adaptive optimization by introducing equipment volatility thresholds (15%, 5%), which significantly improves the robustness of the algorithm in dynamic warehousing scenarios.

[0091] The state prediction unit is used to predict the operating state of the storage equipment;

[0092] The predicting the operating status of the storage equipment includes:

[0093] Based on the pre-processed equipment status data and the first-layer digital twin model, the real-time state equation of the storage equipment is established. The specific formula is:

[0094]

[0095] in, represents the state vector of the storage equipment at time t, f() represents the dynamic relationship function, represents the state vector of the storage equipment at the previous moment, u(t) represents the input signal, p represents the storage equipment parameters, that is, the parameters related to its physical properties, and △t represents the time interval;

[0096] The parameter p represents physical properties related to the system's characteristics, such as mass, friction coefficient, and elastic modulus. These parameters are usually determined through experiments or theoretical derivation during the system modeling phase and determine the system's response to input signals. Overall, the principle of the state equation is based on the time-varying nature of dynamic systems. The system's current state depends not only on its past state but also on external inputs and the system's own characteristics. Through the recursive definition of state, the equation achieves temporal continuity and causality.

[0097] The dynamic relationship function f() represents the dynamic relationship of the storage equipment, describing how to calculate the current state based on the state at the previous moment, the current input signal and the system parameters. This function is usually nonlinear and can capture the complex behavior and interactions of the system.

[0098] In practical applications, this state equation can be used for real-time state prediction, fault diagnosis, and optimized control strategies. By continuously updating the state and making predictions, the health of the system can be monitored in real time, potential faults can be discovered in a timely manner, and the safe and reliable operation of storage equipment can be ensured.

[0099] Update the real-time status of storage equipment. The update formula is:

[0100]

[0101] in, represents the state vector of the storage equipment at time t+△t, Indicates the rate of state change;

[0102] The digital twin model is used on the cloud server to simulate the operating status of the storage equipment in real time. and input signal u(t) to predict the future state of storage equipment.

[0103] The prediction and analysis module is used to monitor warehousing logistics in multiple dimensions, build a fault prediction model, and predict and analyze the failure of warehousing equipment.

[0104] The prediction and analysis module includes: a fault prediction model construction and training unit and a prediction and analysis unit;

[0105] The fault prediction model construction training unit is used to construct a fault prediction model and perform training;

[0106] The fault prediction model is constructed and trained, including:

[0107] Select a machine learning algorithm to build a fault prediction model, using a labeled dataset D train Train the fault prediction model and set the training objective of the model to minimize the loss function. The specific formula of the loss function is:

[0108]

[0109] Where L represents the minimization loss function of the training objective of the fault prediction model, Γ(·) represents the loss function, and xl represents the labeled dataset D train The number of data in y a Denotes the labeled dataset D train The annotation of the data in the data, that is, whether a fault occurs at the time corresponding to the data, Represents the prediction result of the fault prediction model for the ath data;

[0110] The specific formula of the loss function Γ(·) is:

[0111] in, Represents the category weight of the a-th data, which is used to deal with category imbalance. For example, if a certain category has fewer samples, it can be given a higher weight;

[0112] Cross-entropy loss is a common method for measuring the difference between two probability distributions. In classification problems, it evaluates the distance between the predicted distribution output by the model and the true distribution. Weights are introduced to balance the impact of class imbalance. In many practical applications, some categories may have significantly fewer samples than other categories, causing the model to be biased towards the main category. By giving higher weights to minority class samples, the model can attach more importance to these samples during training, thereby improving its prediction performance. The goal of minimizing this loss function is to adjust the model parameters through iterative optimization (such as gradient descent) so that the prediction results are closer to the true label. This method can not only improve the accuracy of the model, but also enhance the ability to identify minority class faults.

[0113] The fault prediction model is trained until the minimization loss function of the model's training objective converges and remains unchanged, and the training is stopped to obtain a trained fault prediction model.

[0114] The prediction and analysis unit uses the trained fault prediction model to predict and analyze the faults of the storage equipment.

[0115] The method of using the trained fault prediction model to predict and analyze storage equipment faults includes:

[0116] Extract the key features from the historical data set D of storage equipment. The key features include statistical features and frequency domain features, and select the feature set F that has the best discriminative power for fault judgment. new , F new ={Ft1,Ft2,...,Ft m}, Ft m It represents the mth feature with the best discriminative ability for fault judgment;

[0117] Specifically, statistical features include mean, standard deviation, maximum, minimum, etc. Frequency domain features extract frequency components through fast Fourier transform (FFT);

[0118] The extracted feature set F new Input the trained storage equipment fault identification model to calculate the storage equipment failure prediction probability. The specific formula is:

[0119]

[0120] Where P(y=c|F new ) indicates that given feature F new The probability of the cth type fault occurring under the condition of , y represents the fault category, The model parameters representing the c-th type of fault are usually weight vectors obtained through training. They represent the relationship between features and fault categories and affect the model's prediction of this category. k represents the index of all possible fault categories, and T represents the total number of fault categories.

[0121] In this embodiment, the principle of the formula is: the formula uses a soft maximization function to convert the linear combination of the model output (i.e. ) is converted to a probability distribution. This function ensures that the sum of the predicted probabilities of all categories is 1, so that the output can be directly interpreted as a probability;

[0122] Model parameters for each class That is, the weight, which is learned through training data and reflects the importance of each feature in different fault categories. The larger the weight, the greater the influence of the feature on identifying the fault of that category;

[0123] By calculating the probability of each category, the model can determine the input feature F new The most likely corresponding fault type is selected as the final prediction result. This formula and its parameters are crucial in fault prediction. By analyzing real-time characteristics and applying corresponding model parameters, potential fault types can be effectively identified and real-time monitoring and maintenance recommendations can be provided.

[0124] Set the fault warning threshold Ψ according to actual needs. If P(y=c|F new)>Ψ, then the alarm is triggered, otherwise, no alarm is triggered.

[0125] In addition, the parts of the above technical solutions provided in the embodiments of the present application that are consistent with the implementation principles of the corresponding technical solutions in the prior art are not described in detail to avoid excessive redundancy.

[0126] The above-described specific embodiments further illustrate the objectives, technical solutions, and beneficial effects of the present invention. It should be understood that the above description is merely a specific embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention shall be included within the scope of protection of the present invention.

Claims

1. The warehouse logistics visualization monitoring system based on digital twin is characterized by: include: Data acquisition module, digital twin construction module, virtual-reality mapping module and prediction analysis module; The data collection module is used to collect various data in the warehouse environment, including: warehouse structure, item location, inventory information, equipment status data and personnel activities, and perform pre-processing; The digital twin construction module constructs a hierarchical digital twin model based on the pre-processed data in the warehouse environment; The virtual-reality mapping module is used to build a real-time mapping system between virtual and real space, and use the digital twin model to simulate and predict the operation of storage equipment; The prediction and analysis module is used to monitor warehousing logistics in multiple dimensions, build a fault prediction model, and predict and analyze the failure of warehousing equipment.

2. The digital twin-based warehouse logistics visualization monitoring system according to claim 1 is characterized in that: The digital twin construction module includes: a first-layer digital twin model construction unit, a second-layer digital twin model construction unit, and a third-layer digital twin model construction unit; The first-layer digital twin model construction unit is used to build a device-level digital twin, model the parameters of the device under different operating conditions, make adjustments based on data, and provide real-time feedback on the operating status of the device; The second-layer digital twin model construction unit is used to build a digital twin at the cargo location level, assigning different color heat values to different cargo locations based on the cargo's residence time and turnover frequency indicators in the cargo location; The third-layer digital twin model construction unit is used to build a system-level digital twin to optimize the overall logistics path planning within the warehouse.

3. The digital twin-based warehouse logistics visualization monitoring system according to claim 1 is characterized in that: The virtual-real mapping module includes: a virtual-real mapping unit and a state prediction unit; The virtual-reality mapping unit is used to synchronize data between the storage equipment and the digital twin model through the Internet of Things technology and to compensate for virtual-reality deviations; The state prediction unit is used to predict the operating state of the storage equipment.

4. The digital twin-based warehouse logistics visualization monitoring system according to claim 3 is characterized in that: The virtual-real deviation compensation includes: The Kalman filter is used to dynamically filter the virtual and real space data to obtain preliminary correction data for the deviation correction of the virtual and real space data; The preliminary correction data is used as the initial population, and the fitness function is defined as the weighted sum of the storage equipment state deviation and the path planning efficiency. Crossover individuals are selected based on the fitness, and the elite individuals are dynamically adjusted. When the storage equipment is in the startup phase and the state volatility is greater than 15%, the top n% of individuals are retained as crossover individuals. When the storage equipment is in the stable phase and the state volatility is less than 5%, m% of crossover individuals are retained, satisfying m>n. Adaptive crossover iteration is used. When the storage equipment is in the startup phase and the state volatility is greater than 15%, the crossover probability is adjusted to 0.

8. When the storage equipment is in the stable phase and the state volatility is less than 5%, the crossover probability is adjusted to 0.

3. If the optimal solution is not updated after 20 iterations or the fitness variance is less than the set threshold for 50 consecutive iterations, part of the population is reset. Randomly select 30% of non-crossover individuals, perform completely random initialization on the selected non-crossover individuals or generate new individuals based on historical data, and perform crossover mutation operations on the randomly initialized individuals or the generated new individuals; When convergence or the maximum evolutionary generation is reached, the optimal deviation correction solution is output.

5. The digital twin-based warehouse logistics visualization monitoring system according to claim 4 is characterized in that: The virtual and real space data is the data of the interaction between warehouse and goods and the digital twin model.

6. The digital twin-based warehouse logistics visualization monitoring system according to claim 4 is characterized in that: The predicting the operating status of the storage equipment includes: Based on the pre-processed equipment status data and the first-layer digital twin model, the real-time state equation of the storage equipment is established; Update the real-time status of the storage equipment according to the status change rate of the storage equipment within △t; The digital twin model is used on the cloud server to simulate the operating status of the storage equipment in real time, and the future status of the storage equipment is predicted by inputting the current status and input signals.

7. The digital twin-based warehouse logistics visualization monitoring system according to claim 1 is characterized in that: The prediction and analysis module includes: a fault prediction model construction and training unit and a prediction and analysis unit; The fault prediction model construction training unit is used to construct a fault prediction model and perform training; The prediction and analysis unit uses the trained fault prediction model to predict and analyze the faults of the storage equipment.

8. The digital twin-based warehouse logistics visualization monitoring system according to claim 7 is characterized in that: The method of using the trained fault prediction model to predict and analyze storage equipment faults includes: Select a machine learning algorithm to build a fault prediction model, using a labeled dataset D train Train the fault prediction model and set the model training objective to minimize the loss function; The fault prediction model is trained until the minimization loss function of the model's training objective converges and remains unchanged, and the training is stopped to obtain a trained fault prediction model.

9. The digital twin-based warehouse logistics visualization monitoring system according to claim 7, characterized in that: The method of using the trained fault prediction model to predict and analyze storage equipment faults includes: Extract the key features from the historical data set D of storage equipment. The key features include statistical features and frequency domain features, and select the feature set F that has the best discriminative power for fault judgment. new , F new ={Ft1,Ft2,...,Ft m }, Ft m It represents the mth feature with the best discriminative ability for fault judgment; The extracted feature set F new Input the trained storage equipment fault identification model to obtain the predicted probability of storage equipment failure; Set the fault warning threshold Ψ according to actual needs. If the predicted probability of storage equipment failure is greater than Ψ, an alarm is triggered; otherwise, no alarm is triggered.

Images