Warehouse management system and method based on active radio frequency identification

Abstract

The invention relates to a warehouse management system and method based on active radio frequency identification, accurate positioning and real-time tracking of warehouse articles are realized through real-time data transmission of active identification and a multi-mode positioning technology, and the problems of inaccurate article positioning, information lag and the like in traditional warehouse management are solved. In combination with an RFID technology and a machine learning algorithm, inventory data are automatically managed, dynamic inventory optimization and intelligent anomaly detection are realized, and errors and time cost of manual operation are reduced. Through the low-power-consumption design and the energy collection technology, the service life of the active identification is prolonged, the maintenance requirement is reduced, the operation cost is reduced, and meanwhile an environment-friendly warehouse management solution is supported. Therefore, a more efficient, more reliable and more intelligent warehouse management system is provided for enterprises, and the complex requirements of modern supply chain management are met.

Description

Technical Field

[0001] The present invention relates to the technical field of warehouse management, and in particular to a warehouse management system and method based on active radio frequency identification. Background Art

[0002] With the rapid development of e-commerce, the importance of warehouse management system (WMS) in supply chain management has become increasingly prominent. Traditional warehouse management usually relies on manual operation and barcode scanning technology. This method has many disadvantages such as information lag, inaccurate inventory and high labor intensity, and cannot meet the needs of modern enterprises for efficient and accurate warehouse management. In order to improve the efficiency and accuracy of warehouse management, radio frequency identification (RFID) technology has gradually been applied to warehouse management systems. RFID technology uses wireless radio frequency signals to automatically identify and track items. It can read the identification information of items without direct contact, greatly improving the automation level of item management. However, traditional passive RFID tags rely on the energy emitted by the reader to work. Although it is suitable for most application scenarios, this passive tag has obvious limitations in some environments that require higher real-time performance and complex environments, such as limited reading distance and slow reading speed. Therefore, the traditional warehouse management system based on RFID tags still has the technical problem of insufficient efficiency of warehouse management in complex warehouse environments. Summary of the invention

[0003] Based on this, it is necessary to provide a warehouse management system based on active radio frequency identification and a warehouse management method based on active radio frequency identification, which can effectively improve the efficiency of warehouse management in complex warehouse environments.

[0004] In order to achieve the above objectives, the embodiments of the present invention adopt the following technical solutions:

[0005] On the one hand, a warehouse management system based on active radio frequency identification is provided, comprising a central processing system, a plurality of RFID readers and writers, and a plurality of active RFID identification modules, wherein each active RFID identification module is respectively attached to each item that needs to be managed in the warehouse, and each RFID reader and writer is respectively installed at each key position in the warehouse, and each RFID reader and writer is respectively connected to the central processing system for communication and is used to receive item information sent by each active RFID identification module; the item information includes a unique identification code, location information, status information, basic item data, and environmental parameters of the storage area where the item is located, and the key positions include each storage area, entrance and exit, and shelf of the warehouse; each active RFID identification module has a built-in power supply, and is used to actively send item information periodically or upon receiving a read instruction sent by the RFID reader and writer;

[0006] The central processing system is used to receive, process, store and analyze the item information sent back by each RFID reader, and update the location and status of the item corresponding to the item information in real time. The central processing system uses the LSTM model to predict the future inventory demand results of the items based on the historical inventory data of the items in the warehouse, and outputs inventory adjustment suggestions based on the predicted inventory demand results. The central processing system uses an anomaly detection algorithm to analyze the inventory data of the items in the warehouse and identify abnormal behaviors, and issue an early warning when an anomaly is identified; among them, the central processing system is configured with an item management model and a warehouse layout model. The item management model is used to manage the inventory status, location, movement trajectory, service life of the items in the warehouse in real time and monitor the storage environment of the items. The warehouse layout model is used to manage the distribution of each storage area in the warehouse, the flow of items in each storage area and the inventory data through a graphical interface.

[0007] On the other hand, a warehouse management method based on active radio frequency identification is also provided, which is applied to the above-mentioned warehouse management system based on active radio frequency identification. The warehouse management method based on active radio frequency identification comprises the following steps:

[0008] Install RFID readers at key locations in the warehouse and attach active RFID identification modules to items;

[0009] Communicate each RFID reader to the central processing system and initialize the system; initializing the system includes initializing the database of the central processing system and entering basic data of the items;

[0010] The central processing system obtains the item information sent by each active RFID identification module received and returned by each RFID reader;

[0011] The central processing system processes, stores and analyzes item information, and updates the location and status of items corresponding to the item information in real time;

[0012] The central processing system uses the LSTM model to predict the future inventory demand results of items based on the historical inventory data of items in the warehouse, and outputs inventory adjustment suggestions based on the predicted inventory demand results;

[0013] The central processing system uses anomaly detection algorithms to analyze inventory data of items in the warehouse and identify abnormal behavior, and issues warnings when anomalies are identified.

[0014] One of the above technical solutions has the following advantages and beneficial effects:

[0015] The above-mentioned warehouse management system and method based on active radio frequency identification have the ability to actively send signals by introducing an active RFID identification module, which significantly improves the real-time and accuracy of item tracking. It is especially suitable for complex warehousing environments. At the same time, it predicts and analyzes inventory data through a machine learning algorithm (LSTM), automatically optimizes inventory levels, reduces inventory backlogs or shortages, and improves supply chain efficiency. It also applies an anomaly detection algorithm (Isolation Forest) to monitor inventory data in real time, automatically identify and warn of abnormal behaviors, ensure inventory security and timeliness of management. At the same time, it combines RFID with other positioning technologies (such as Wi-Fi and UWB) to achieve higher-precision item positioning, solving the shortcomings of traditional technologies in precise positioning in complex warehousing environments.

[0016] Compared with traditional technologies, the above solution realizes real-time tracking of items, automated inventory management, abnormal state detection and early warning, and intelligent analysis and optimization of inventory data in the warehouse management system through an active identification system based on radio frequency identification (RFID) technology, thereby improving the efficiency, accuracy and response speed of warehouse management, reducing manual intervention and reducing operating costs. For example, through the real-time data transmission and multi-mode positioning technology of active identification, accurate positioning and real-time tracking of stored items can be achieved, solving the problems of inaccurate item positioning and information lag in traditional warehouse management. Combining RFID technology with machine learning algorithms, inventory data can be automatically managed, dynamic inventory optimization and intelligent anomaly detection can be achieved, and errors and time costs of manual operations can be reduced. Through low-power design and energy harvesting technology, the service life of active identification can be extended, maintenance requirements can be reduced, operating costs can be reduced, and green and environmentally friendly warehouse management solutions can be supported. In this way, enterprises are provided with a more efficient, reliable and intelligent warehouse management system to meet the complex needs of modern supply chain management. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The drawings constituting a part of this application are used to provide a further understanding of the present invention. The illustrative embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations on the present invention.

[0018] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0019] Figure 1 Schematic diagram of the overall structure of a warehouse management system based on active radio frequency identification in one embodiment;

[0020] Figure 2A schematic diagram of a core application process of a warehouse management system based on active radio frequency identification in one embodiment;

[0021] Figure 3 The figure is a flowchart of a warehouse management method based on active radio frequency identification in one embodiment. DETAILED DESCRIPTION

[0022] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific embodiments of the present invention are described in detail below in conjunction with the accompanying drawings. In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention can be implemented in many other ways different from those described herein, and those skilled in the art can make similar improvements without violating the connotation of the present invention, so the present invention is not limited by the specific embodiments disclosed below.

[0023] The following will describe the implementation of the present invention in detail with reference to the accompanying drawings in the embodiment diagram of the present invention.

[0024] In traditional technology, some RFID-based warehouse management solutions have been proposed and applied, such as warehouse management systems based on passive RFID tags, where passive RFID tags are widely used, especially in inventory counting and item in and out management, using RFID readers to scan RFID tags to obtain item information. Although this system can improve the efficiency of inventory management to a certain extent, due to the limited reading distance of passive RFID tags and reliance on the excitation signal of the RFID reader, it lacks real-time performance and flexibility.

[0025] Another example is hybrid RFID systems, which try to combine passive RFID tags with active tags to take advantage of the advantages of both. For example, active tags are used on high-value or important items to achieve higher security and real-time tracking, while lower-cost passive RFID tags are used on other items. This hybrid system improves the flexibility of the system to a certain extent, but also increases the complexity and cost of management.

[0026] Or, for example, systems that combine RFID technology with other positioning technologies, such as combining RFID, Bluetooth, Wi-Fi, and UWB (ultra-wideband) technologies to enhance positioning capabilities. Such systems achieve higher-precision item tracking through multi-mode positioning technology, but they also face problems of high cost and complex technical integration. Another example is the system that combines the Internet of Things (IoT) with RFID. The development of the Internet of Things has promoted the further application of RFID technology in warehouse management. These systems achieve real-time data transmission and remote monitoring by connecting RFID tags to the Internet of Things platform. Such systems enhance the interconnectivity of data, but they rely on the construction and maintenance of the Internet of Things infrastructure.

[0027] In this context, active RFID tags have begun to be used in the field of warehouse management. Active RFID tags have their own power supply, can actively send signals, have a larger reading range and higher real-time performance, and are particularly suitable for scenarios that require frequent updates and high-precision positioning. However, how to effectively use active RFID tags in complex warehouse environments that require higher real-time performance to achieve accurate tracking and intelligent management of items is still a problem that needs to be solved.

[0028] In one embodiment, see Figure 1 , a warehouse management system 100 based on active radio frequency identification is provided, including a central processing system 11, multiple RFID readers 13 and multiple active RFID identification modules 15. Each active RFID identification module 15 is attached to each item that needs to be managed in the warehouse. Each RFID reader 13 is installed in each key position in the warehouse, and each RFID reader 13 is respectively connected to the central processing system 11 for communication and is used to receive the item information sent by each active RFID identification module 15. The item information includes the unique identification code, location information, status information, basic data of the item and the environmental parameters of the storage area where the item is located. The key positions include the storage areas, entrances and exits and shelves of the warehouse. Each active RFID identification module 15 has a built-in power supply, which is used to actively send item information periodically or upon receiving a read instruction sent by the RFID reader 13.

[0029] The central processing system 11 is used to receive, process, store and analyze the item information sent back by each RFID reader 13, and update the location and status of the item corresponding to the item information in real time. The central processing system 11 uses the LSTM model to predict the inventory demand results of the items in the future based on the historical inventory data of the items in the warehouse, and outputs inventory adjustment suggestions based on the predicted inventory demand results. The central processing system 11 uses an anomaly detection algorithm to analyze the inventory data of the items in the warehouse and identify abnormal behaviors, and issue an early warning when an abnormality is identified. Among them, the central processing system 11 is configured with an item management model and a warehouse layout model. The item management model is used to manage the inventory status, location, movement trajectory, service life of the items in the warehouse in real time, and monitor the storage environment of the items. The warehouse layout model is used to manage the distribution of each storage area in the warehouse, the flow of items in each storage area, and the inventory data through a graphical interface.

[0030] Understandably, Figure 1Where N and M are positive integers not less than 2, solid lines indicate that the modules are connected by wired communication or wireless communication, and dotted lines indicate that the modules are connected by wireless communication. The RFID reader 13 can use various existing RFID readers, and the active RFID identification module 15 can use various existing active RFID tags. The central processing system 11 can be an MCU unit or a CPU unit, or a local warehouse management server system or a computer host.

[0031] The above-mentioned warehouse management system 100 based on active radio frequency identification has the ability to actively send signals by introducing an active RFID identification module 15, which significantly improves the real-time and accuracy of item tracking, and is particularly suitable for complex warehousing environments. At the same time, it predicts and analyzes inventory data through a machine learning algorithm (LSTM), automatically optimizes inventory levels, reduces inventory backlogs or shortages, and improves supply chain efficiency. It also applies an anomaly detection algorithm (Isolation Forest) to monitor inventory data in real time, automatically identifies and warns of abnormal behaviors, and ensures inventory security and timeliness of management. At the same time, it combines RFID with other positioning technologies (such as Wi-Fi and UWB) to achieve higher-precision item positioning, solving the shortcomings of traditional technologies in precise positioning in complex warehousing environments.

[0032] Compared with traditional technologies, the above solution realizes real-time tracking of items, automated inventory management, abnormal state detection and early warning, and intelligent analysis and optimization of inventory data in the warehouse management system through an active identification system based on radio frequency identification (RFID) technology, thereby improving the efficiency, accuracy and response speed of warehouse management, reducing manual intervention and reducing operating costs. For example, through the real-time data transmission and multi-mode positioning technology of active identification, accurate positioning and real-time tracking of stored items can be achieved, solving the problems of inaccurate item positioning and information lag in traditional warehouse management. Combining RFID technology with machine learning algorithms, inventory data can be automatically managed, dynamic inventory optimization and intelligent anomaly detection can be achieved, and errors and time costs of manual operations can be reduced. Through low-power design and energy harvesting technology, the service life of active identification can be extended, maintenance requirements can be reduced, operating costs can be reduced, and green and environmentally friendly warehouse management solutions can be supported. In this way, enterprises are provided with a more efficient, reliable and intelligent warehouse management system to meet the complex needs of modern supply chain management.

[0033] It should be noted that in practical applications, Figure 2As shown, first of all, equipment deployment and system initialization can be carried out: RFID readers 13 are installed at key locations in the warehouse (such as storage areas, entrances and exits, and shelves, etc.), and these RFID readers 13 will continuously monitor the RFID identification activities in the area. In the environment inside the warehouse, the number and layout of the RFID readers 13 can be adjusted and optimized accordingly according to the application needs of the actual scenario to ensure that all areas have signal coverage without dead spots. Each stored item in the warehouse is attached with an active RFID identification module 15, which is configured with a unique identification code of the item, and has built-in sensors for monitoring environmental data (such as temperature and humidity, etc.). The active RFID identification module 15 will be activated during initialization, and will periodically send its unique identification code and related data to nearby RFID readers 13 through wireless communication.

[0034] Among them, the active RFID identification module 15 carries an active RFID identification and has a built-in power supply, and can actively send signals (including identification information of the item) regularly or according to external trigger conditions. The identification information is used to record data such as the unique identification code, location information and status information of the item. The active RFID identification module 15 can actively send data regularly or according to external trigger conditions (such as the movement of the item). The RFID reader 13 receives data in real time, and the received data may include information such as the unique identification code, timestamp, location information and status information of the item.

[0035] In some embodiments, although the active RFID identification module 15 can actively send signals, it also faces high energy consumption requirements. In order to solve the problem of insufficient power of the active RFID identification module 15, the central processing system 11 has a built-in energy management module, which combines wireless charging technology or solar charging technology to provide power support for the active RFID identification module 15.

[0036] It is understood that the active RFID identification module 15 adopts a low power design to extend battery life and reduce replacement frequency. In some embodiments, the active RFID identification module 15 and the RFID reader 13 can use energy collection technology (such as solar power supply or wireless charging) to further reduce energy consumption.

[0037] In some embodiments, when the power of the active RFID identification module 15 is low, the central processing system 11 will automatically issue a warning to remind the warehouse manager to maintain or replace the active RFID identification module 15. The central processing system 11 will track the battery status of each active RFID identification module 15 and optimize its charging time to ensure that the active RFID identification module 15 is always in working condition.

[0038] The signal range of the antenna module of the RFID reader 13 covers the entire warehouse area, ensuring seamless signal reception. The RFID reader 13 is used to receive signals from active RFID tags and transmit the collected data to the central processing system 11. The data collected by the RFID reader 13 can be transmitted to the central processing system 11 in real time by wired communication or wireless communication. The stability and security of the signal are ensured during data transmission.

[0039] The central processing system 11 is electrically connected to the warehouse management system (WMS) to achieve data synchronization and intercommunication, and the system database is initialized to store inventory information, item attributes and location information, etc. At this stage, it is also necessary to enter the basic data of the stored items, which may include the name, category, quantity, storage time and expected delivery time of the items. An item management model is established in the central processing system 11, and each item is represented by its unique identification code. The item management model may include components such as inventory status, location, movement trajectory and service life. The sensor data of the active RFID identification module 15 (used to monitor environmental parameters) is also included in the item management model for real-time monitoring of the item storage environment.

[0040] The central processing system 11 is configured with a warehouse layout model for managing the distribution of each storage area, the flow of goods in each storage area and inventory data. The warehouse layout model can be configured through the system's graphical interface to ensure that the warehouse manager can clearly view the distribution and status of goods in each area.

[0041] The central processing system 11 is responsible for receiving, processing, storing and analyzing data. A distributed central processing system 11 can be used to ensure high concurrent processing capabilities and high availability. The system can pre-process the collected data, such as cleaning, removing duplicate, erroneous and missing data. After the system pre-processes the data, it can normalize and format the data to ensure data consistency. The processed data can be stored in the database of the central processing system 11, supporting the storage of structured and unstructured data.

[0042] In some embodiments, the central processing system 11 calculates the real-time and accurate location of the object using the signal strength of the RFID reader / writer 13 and the location of the active RFID identification module 15 .

[0043] It is understood that the central processing system 11 can calculate the precise location of the item based on the signal strength and identification location of the RFID reader 13, for example, by using multi-mode positioning technology (such as RFID combined with Wi-Fi) to improve positioning accuracy. Then, the central processing system 11 updates the location and status of the item in real time, and generates a visual warehouse layout diagram to display the real-time and precise location of the item.

[0044] The central processing system 11 can also use machine learning algorithms (such as long short-term memory network, Long Short-Term Memory, i.e. LSTM model) to analyze historical inventory data and predict the future inventory demand results of items, and then automatically adjust the inventory level according to the predicted inventory demand results to avoid excessive or insufficient inventory.

[0045] It is understood that the central processing system 11 can also perform inventory optimization and analysis. Specifically, the historical inventory data is analyzed to identify inventory fluctuations and item demand trends, and the inventory is automatically optimized based on the analysis results. For example, the central processing system 11 can predict the peak demand period of a certain type of item in the future by analyzing the historical inventory data, and then give suggestions for early replenishment. The central processing system 11 can use machine learning algorithms (such as LSTM models) to analyze and predict historical inventory data, optimize inventory levels, and provide replenishment suggestions.

[0046] The LSTM model can control the information flow of inventory data through three gates (input gate, forget gate, and output gate):

[0047] Input Gate: determines how much current memory to update, that is, determines what information to include in the LSTM model at the current moment. It has the following two steps:

[0048] First, the sigmoid function is used to determine the part that needs to be updated:

[0049] i t =σ(W i ·[h t-1 ,x t ]+b i )

[0050] Secondly, a new candidate memory for the current moment is generated through the tanh function:

[0051] C t =tanh(W C ·[h t-1 ,x t ]+b C )

[0052] In the central processing system 11, the information that the input gate can process includes: inventory (real-time inventory of current items), replenishment records (data of past replenishments), sales fluctuations (sales history of items or demand forecasts), and goods status (including item loss rate, inventory turnover rate, etc.). The input gate is responsible for determining the impact of the new information incorporated into the LSTM model on the current forecast. Through the activation of the input gate, the LSTM model can determine which information is most important for future inventory changes based on the inventory and sales information at the current moment. For example, if the recent sales fluctuations are more severe, the input gate will give more weight to the demand forecast data. After being processed by the input gate, the input related information is integrated into the memory unit, laying the foundation for subsequent status updates and predictions.

[0053] Forget Gate: determines how much previous memory should be forgotten. Its formula is:

[0054] f t =σ(W f ·[h t-1 ,x t ]+b f )

[0055] Among them, i t is the output of the input gate, W i and b i are the weight matrix and bias term of the input gate, respectively, and f t is the output of the forget gate, σ is the sigmoid activation function, W f and b f are the weight matrix and bias term of the forget gate, h t-1 is the hidden state at the previous moment, x t is the input at the current moment.

[0056] Memory update: Through the joint action of the forget gate and the input gate, the memory state of the current moment is updated:

[0057] C t =f t ·C t-1 +i t ·C t

[0058] Among them, C t is the memory state at the current moment, C t-1 is the memory state of the previous moment,

[0059] Specifically, the forget gate controls which historical information needs to be "forgotten" or ignored. In item management, the forget gate can determine historical inventory data (for older inventory data, the forget gate can choose to reduce its importance) and outdated demand patterns (that is, if market demand has changed significantly, the forget gate can choose to ignore the old demand pattern and focus on new data instead). The forget gate plays a role in filtering invalid information when processing complex inventory data. Especially in inventory forecasting, some old data may no longer have reference value or there may be seasonal demand changes. The forget gate can automatically adjust the weights to reduce the interference of these old data on future forecasts. After the operation of the forget gate, the LSTM model will delete data that is no longer important and retain information that is still useful for future forecasts.

[0060] Output Gate: Determines the output content of the LSTM unit in the LSTM model. It first controls the output part through the sigmoid function:

[0061] o t =σ(W o ·[h t-1 ,x t ]+b o )

[0062] Then combine the current memory state to generate the final output:

[0063] h t =o t tanh(C t )

[0064] Among them, t is the output of the LSTM unit at the current moment, W o and b o are the weight matrix and bias term of the output gate, h t is the hidden state or output at the current moment.

[0065] Specifically, the output gate determines the output of the unit of the LSTM model at the current moment, and is usually used as a direct source of the prediction results. In warehouse management, the output gate can process inventory forecast results (i.e., predicting inventory change trends in the next few days or weeks) and replenishment needs (such as predicting the amount and time of replenishment required). The output gate can generate the inventory forecast for the next moment based on the current inventory data, demand changes, and the comprehensive processing results of the forget gate and input gate. This is of great significance for inventory optimization, reducing warehousing costs, and preventing inventory shortages (or excesses). The output gate ensures that at each time step, the LSTM model outputs the most relevant prediction information. The output gate ultimately determines the prediction results, such as the inventory level, sales trend, or replenishment demand at the next point in time.

[0066] In the above LSTM model, the input gate plays a significant role, especially in real-time inventory management. When there is a new order demand or replenishment information, the input gate will dynamically adjust the memory state of the LSTM model so that the model can capture the latest market demand and inventory dynamics. This is particularly important when dealing with rapidly changing market demand. Since the data in the warehouse is time-dependent, but not all historical data can play a role in future predictions, the design of the above forget gate enables the LSTM model to actively "forget" those data that are no longer important. Especially in the case of a short inventory turnover cycle or large fluctuations in market demand, this mechanism can effectively prevent the impact of outdated information on the prediction results. In the application of inventory forecasting and anomaly detection, the output gate is responsible for generating prediction results and deciding whether to issue an alarm, such as when to replenish or avoid insufficient inventory. The output gate can combine multiple input factors (such as sales volume and demand changes) to generate highly accurate predictions, supporting decision makers to adjust inventory strategies in real time.

[0067] In some embodiments, the central processing system 11 is also used to calculate the optimal inventory quantity of items in the warehouse and automatically generate replenishment suggestions based on the current inventory level of items in the warehouse, predicted inventory demand results and procurement cycle through a linear regression model; the central processing system 11 is configured with an automatic ordering function.

[0068] It can be understood that the central processing system 11 can also apply a linear regression model to calculate the optimal inventory quantity based on factors such as the current inventory level of the item, the predicted inventory demand results and the procurement cycle, and then the central processing system 11 can automatically generate replenishment suggestions and support automatic ordering functions. For example, the central processing system 11 generates a replenishment order for the item that is currently in need of timely replenishment as analyzed. The information in the replenishment order includes information such as the item name, storage area, recommended replenishment quantity, recommended replenishment date, and estimated order price. Then, the central processing system 11 pushes the generated replenishment order to the terminal used by the warehouse manager via SMS, email, order push of the bound procurement platform, or other forms, so that the warehouse manager can click to confirm the replenishment and then automatically push the replenishment order to the manufacturer of the item, thereby realizing the automatic ordering function.

[0069] In some embodiments, the central processing system 11 may also use an anomaly detection algorithm (IsolationForest) to analyze inventory data, identify abnormal behavior (such as a sudden decrease in inventory or a large number of items being moved), and then automatically trigger an early warning mechanism to issue an early warning based on the identification results. When an abnormality is identified, the central processing system 11 may issue an early warning message via SMS, email, or mobile application to notify relevant management personnel. The early warning information may include the type of abnormality, the affected items, and recommended treatment measures.

[0070] It can be understood that the central processing system 11 analyzes the real-time inventory data to monitor inventory changes. When an abnormal situation occurs (such as unauthorized movement of items or abnormal outbound delivery, etc.), the central processing system 11 will immediately identify and issue an alarm. Abnormal situations can specifically include items leaving the designated storage area, sudden inventory reduction, and sensor data abnormalities (such as excessive temperature or abnormal humidity). Specifically, the central processing system 11 also has an abnormality detection function, which can use an abnormality detection algorithm (Isolation Forest) to identify abnormal behavior in the inventory (such as unauthorized withdrawal and loss, etc.) and automatically trigger an alarm.

[0071] The mathematical model of the anomaly detection algorithm is mainly based on the calculation of path length and anomaly score. Specifically, the first is the construction of the tree and the path length: the tree is constructed by randomly selecting features and cutting points, and the path length represents the depth of a sample point to reach the leaf node. Assume that the depth of the tree is h(x), that is, the number of cuts required for sample x to be isolated in an isolated tree. Among them, sample x represents a specific item-related data, which may include the following feature data points: inventory (i.e. the current inventory quantity of the item), storage cycle (i.e. the retention time of the item in the warehouse), in-and-out records (i.e. the historical out-and-in data of each item), sales history (i.e. the sales fluctuation of the item over a period of time), replenishment cycle (i.e. the frequency of item replenishment and the time of the last replenishment) and item loss rate (considering that some items may have a certain amount of natural loss, the sample contains the historical loss of each item). These feature data points are input as samples into the mathematical model of the anomaly detection algorithm, which can help the model understand the dynamic characteristics of the item and improve the accuracy of inventory prediction and anomaly detection through long-term analysis of these data.

[0072] For a stock dataset of size n, the expected value of the path length can be expressed as:

[0073]

[0074] Where H(i) is the i-th harmonic number (i is an integer from 1 to n), which can be expressed as:

[0075]

[0076] Formula c(n) represents the expected path length for normalizing the tree. In Isolation Forest, the path length reflects the ease with which a sample is isolated in the tree structure. For the item management in this embodiment, the path length can reflect: Inventory abnormality. If the inventory of an item is significantly different from other items (for example, the inventory suddenly decreases or increases sharply), it will be isolated earlier in the tree structure of Isolation Forest, and the path length will be shorter; this can effectively detect inventory abnormalities. Sales abnormality. If the sales of an item are abnormal (for example, sales suddenly surge or decline), Isolation Forest can quickly identify abnormal sales patterns by analyzing the path lengths of these data points. Replenishment cycle abnormality. When the replenishment frequency of an item is obviously inconsistent with historical data (for example, an item is rarely replenished in normal times but suddenly replenished frequently), the model can detect this abnormality through a shorter path length.

[0077] Then perform anomaly score calculation: The ultimate goal is to calculate the anomaly score of a sample. The anomaly score is based on the average path length E(h(x)) of the sample in multiple isolated trees, that is, the average depth of the sample isolated in multiple trees.

[0078] The anomaly score S(x,n) of a sample is calculated using the following formula:

[0079]

[0080] Where E(h(x)) is the average path length of sample x in multiple trees.

[0081] When the anomaly score of a sample is close to 1, it means that the sample is easy to isolate and is an outlier; when the score is close to 0, it means that the sample is more similar to other samples and is a normal point.

[0082] Once the central processing system 11 detects an abnormal situation using the anomaly score, the central processing system 11 will automatically send a notification to the warehouse manager. The notification can be sent in a variety of ways, such as but not limited to SMS, email or mobile phone application reminders. The warehouse manager can view the abnormal details through the system interface of the central processing system 11 and take corresponding treatment measures in a timely manner.

[0083] The special role of the above-mentioned anomaly detection algorithm in this embodiment is not only anomaly detection, but also can reflect abnormal fluctuations in item inventory and sales trends through the path length of the data, thereby providing early warning for inventory management. Compared with traditional algorithms, the above-mentioned anomaly detection algorithm is not simply misappropriated, but by refining the association between path length and item data, it realizes accurate detection of inventory anomalies and replenishment frequency. Its specific special roles can be summarized as follows: Early anomaly warning, by identifying abnormal patterns in inventory, sales and replenishment, the algorithm can issue an early warning before the problem occurs, helping warehouse managers to take action in advance (such as adjusting inventory and optimizing replenishment strategies). Dynamic inventory adjustment, according to changes in item inventory and sales, the above-mentioned anomaly detection algorithm can automatically adjust the monitoring standards, especially when market demand fluctuates violently, this adjustment can enhance the flexibility and adaptability of prediction.

[0084] In addition, the central processing system 11 will also record all abnormal detection and processing processes and generate abnormal processing logs to facilitate future review or optimization of processes. For abnormal situations that require further follow-up, the central processing system 11 can set up a multi-level alarm mechanism to ensure that abnormal problems are resolved in a timely manner.

[0085] Specifically, when an item enters the warehouse, the active RFID identification module 15 will be automatically identified by the RFID reader 13 and the basic information of the item (such as item number, item type and quantity, etc.) will be sent to the central processing system 11. After receiving the basic information of the item, the central processing system 11 will allocate the item to the corresponding storage location and update the data in the database. The item management model will automatically allocate the most suitable storage area for the item based on the category, demand and storage conditions of the item. After the item enters the designated storage area, the active RFID identification module 15 will regularly send location information and environmental data, and the central processing system 11 will update the inventory status of the item in real time to ensure that the inventory data is consistent with the actual situation.

[0086] When items need to be shipped out of the warehouse, the warehouse manager selects the items that need to be shipped out through the central processing system 11, and issues the shipping instructions for the items through the central processing system 11, so that the active RFID identification module 15 and the RFID reader 13 can locate the position of the items in the warehouse in real time, ensuring that the warehouse manager can quickly find the items that need to be shipped out. When the items are taken out, the signal of the active RFID identification module 15 will be automatically updated to the central processing system 11, so that the time, location and quantity of the items leaving the warehouse will be synchronously updated to the system database. After the items are shipped out of the warehouse, the central processing system 11 will automatically update the inventory, display the current inventory status and generate the corresponding shipping report for the warehouse manager to review.

[0087] In some embodiments, the system can monitor the inventory status in real time. Specifically, the central processing system 11 monitors the status of the active RFID identification module 15 in real time through the RFID reader 13, obtains the location information and environmental parameters of the items, and ensures that the location and environmental parameters of each item are under monitoring. Warehouse managers can view the detailed information of each item, including its inventory status, location information, and environmental conditions, through the visual interface on the system.

[0088] The central processing system 11 can also automatically generate inventory reports based on the storage time and inventory of items to help warehouse managers understand the current inventory status. For items in short supply, the central processing system 11 will issue replenishment reminders in advance to ensure that the warehouse has sufficient items.

[0089] In one embodiment, Figure 3 As shown, a warehouse management method based on active radio frequency identification is also provided, which is applied to the above-mentioned warehouse management system 100 based on active radio frequency identification. The warehouse management method based on active radio frequency identification may include the following processing steps S10 to S20:

[0090] S10, installing RFID readers at key locations in the warehouse and attaching active RFID identification modules to items;

[0091] S12, communicating each RFID reader to the central processing system and initializing the system; initializing the system includes initializing the database of the central processing system and entering basic data of the items;

[0092] S14, the central processing system obtains the item information sent by each active RFID identification module and received and returned by each RFID reader;

[0093] S16, the central processing system processes, stores and analyzes the item information, and updates the location and status of the item corresponding to the item information in real time;

[0094] S18, the central processing system uses the LSTM model to predict the future inventory demand results of items based on the historical inventory data of items in the warehouse, and outputs inventory adjustment suggestions based on the predicted inventory demand results;

[0095] S20, the central processing system uses an anomaly detection algorithm to analyze the inventory data of items in the warehouse and identify abnormal behavior, and issues an early warning when an anomaly is identified.

[0096] The above-mentioned warehouse management method based on active radio frequency identification realizes real-time tracking of items, automated inventory management, abnormal state detection and early warning, and intelligent analysis and optimization of inventory data in the warehouse management system through an active identification system based on radio frequency identification (RFID) technology, thereby improving the efficiency, accuracy and response speed of warehouse management, reducing manual intervention, and reducing operating costs. For example, through the real-time data transmission and multi-mode positioning technology of active identification, accurate positioning and real-time tracking of stored items are achieved, solving the problems of inaccurate item positioning and information lag in traditional warehouse management. Combining RFID technology with machine learning algorithms, inventory data is automatically managed, dynamic inventory optimization and intelligent anomaly detection are achieved, and errors and time costs of manual operations are reduced. Through low-power design and energy harvesting technology, the service life of active identification is extended, maintenance requirements are reduced, operating costs are reduced, and green and environmentally friendly warehouse management solutions are supported. In this way, a more efficient, reliable and intelligent warehouse management system is provided for enterprises to meet the complex needs of modern supply chain management.

[0097] In one embodiment, the above-mentioned warehouse management method based on active radio frequency identification may further include the steps of:

[0098] After the central processing system obtains the signal strength of the RFID reader and the location of the active RFID identification module, it calculates the real-time and accurate location of the object and updates it to the location information of the object.

[0099] In one embodiment, the above-mentioned warehouse management method based on active radio frequency identification may further include the steps of:

[0100] After the central processing system calculates the optimal inventory of items in the warehouse and automatically generates replenishment suggestions based on the current inventory level of items in the warehouse, the predicted inventory demand results and the procurement cycle through a linear regression model, it sends the automatically generated item replenishment order to the user terminal for confirmation. The user terminal is, for example, but not limited to, a mobile phone, a tablet computer or a computer of a warehouse management agent.

[0101] It can be understood that for the specific limitations of the warehouse management method based on active radio frequency identification, reference can be made to the corresponding limitations of the various embodiments of the warehouse management system 100 based on active radio frequency identification described above, and no further details will be given here.

[0102] It should be understood that although Figure 3 The steps in the method are shown in sequence as indicated by the arrows, but the steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified in this document, there is no strict order restriction for the execution of the steps, and the steps can be executed in other orders. Figure 3At least part of the steps may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily executed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequentially, but can be executed in turn or alternately with other steps or at least part of the sub-steps or stages of other steps.

[0103] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0104] The above embodiments only express several implementation methods of the present invention, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for those of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present invention, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the patent of the present invention shall be subject to the attached claims.

[0105] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least one of the features. In the description of the present invention, the meaning of "plurality" is at least two, such as two, three, etc., unless otherwise clearly and specifically defined.

[0106] In the present invention, unless otherwise clearly defined and specified, the terms "connected" and "connection" should be understood in a broad sense, for example, it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium; it can be the internal connection of two elements or the interaction relationship between two elements, unless otherwise clearly defined. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

Claims

1. A warehouse management system based on active radio frequency identification, characterized in that: It includes a central processing system, multiple RFID readers and writers, and multiple active RFID identification modules. Each of the active RFID identification modules is attached to each item that needs to be stored and managed. Each of the RFID readers and writers is installed at each key position in the warehouse. Each of the RFID readers and writers is connected to the central processing system for communication and is used to receive item information sent by each of the active RFID identification modules. The item information includes the unique identification code, location information, status information, basic item data, and environmental parameters of the storage area where the item is located. The key positions include storage areas, entrances and exits, and shelves in the warehouse. Each of the active RFID identification modules has a built-in power supply, which is used to actively send the item information periodically or upon receiving a read instruction sent by the RFID reader and writer. The central processing system is used to receive, process, store and analyze the item information returned by each of the RFID readers, and update the location and status of the items corresponding to the item information in real time. The central processing system uses an LSTM model to predict the future inventory demand results of items based on the historical inventory data of items in the warehouse, and outputs inventory adjustment suggestions based on the predicted inventory demand results. The central processing system uses an anomaly detection algorithm to analyze the inventory data of items in the warehouse and identify abnormal behaviors, and issue an early warning when an anomaly is identified; wherein, the central processing system is configured with an item management model and a warehouse layout model. The item management model is used to manage the inventory status, location, movement trajectory, service life of items in the warehouse in real time and monitor the storage environment of items. The warehouse layout model is used to manage the distribution of each storage area in the warehouse, the flow of items in each storage area and inventory data through a graphical interface.

2. The warehouse management system based on active radio frequency identification according to claim 1 is characterized in that: The central processing system calculates the real-time and accurate location of the object using the signal strength of the RFID reader and the location of the active RFID identification module.

3. The warehouse management system based on active radio frequency identification according to claim 1 or 2, characterized in that: The central processing system is also used to calculate the optimal inventory quantity of the items in the warehouse and automatically generate replenishment suggestions based on the current inventory level of the items in the warehouse, the predicted inventory demand results and the procurement cycle through a linear regression model; The central processing system is equipped with an automatic order placing function.

4. A warehouse management method based on active radio frequency identification, characterized in that: Applied to the warehouse management system based on active radio frequency identification as described in any one of claims 1 to 3, the warehouse management method based on active radio frequency identification comprises the steps of: Install RFID readers at key locations in the warehouse and attach active RFID identification modules to items; Communicate each of the RFID readers to the central processing system and initialize the system; initializing the system includes initializing the database of the central processing system and entering basic data of items; The central processing system obtains the item information sent by each active RFID identification module and received and returned by each RFID reader; The central processing system processes, stores and analyzes the item information, and updates the location and status of the item corresponding to the item information in real time; The central processing system uses the LSTM model to predict the future inventory demand results of items based on the historical inventory data of items in the warehouse, and outputs inventory adjustment suggestions based on the predicted inventory demand results; The central processing system uses an anomaly detection algorithm to analyze the inventory data of items in the warehouse and identify abnormal behavior, and issues an early warning when an anomaly is identified.

5. The warehouse management method based on active radio frequency identification according to claim 4 is characterized in that: Also includes the steps: After obtaining the signal strength of the RFID reader and the location of the active RFID identification module, the central processing system calculates the real-time and accurate location of the object and updates the location information of the object.

6. The warehouse management method based on active radio frequency identification according to claim 4 is characterized in that: Also includes the steps: After the central processing system calculates the optimal inventory quantity of the items in the warehouse and automatically generates replenishment suggestions based on the current inventory level of the items in the warehouse, the predicted inventory demand results and the procurement cycle through a linear regression model, the automatically generated item replenishment order is sent to the user terminal for confirmation.

Images