Access control method and system based on ultrahigh frequency RFID technology

Abstract

The invention relates to the field of access control, in particular to an access control method and system based on the ultrahigh frequency RFID technology, and the method comprises the following steps: S1, setting a multi-antenna array through distributed layout according to the size, shape and personnel flow condition of an access control area; s2, according to a Hyperledger Fabric block chain platform, nodes are deployed at different positions through a distributed node architecture; s3, the artificial intelligence model is trained by collecting data in the access control system, the trained artificial intelligence model analyzes and detects the data in the access control system in real time, and abnormal behaviors and events are recognized; and S4, displaying each data in the access control system on a management platform in a visual mode, and providing data analysis and optimization management according to an authority management mechanism. The safety of the access control system can be improved, the reliability of the access control system is improved, the intelligent level of the access control system is enhanced, and the management efficiency of the access control system is improved.

Description

Technical Field

[0001] The present invention relates to the field of access control, and in particular to an access control method and system based on ultra-high frequency RFID technology. Background Art

[0002] With the continuous development of science and technology, the importance of access control systems in various fields has become increasingly prominent. Traditional access control systems mainly use keys, passwords, cards and other methods for identity authentication, but these methods have problems such as easy loss, easy cracking, and inconvenient management. The emergence of UHF RFID technology has brought new solutions to access control systems. UHF RFID has the advantages of long-distance identification, fast reading, and simultaneous identification of multiple tags, which can effectively improve the efficiency and security of access control systems. However, traditional UHF RFID technology also has some limitations, such as data may be tampered with, signal interference, and inability to effectively detect abnormal behavior. Summary of the invention

[0003] To solve the above problems, the present invention provides an access control method and system based on ultra-high frequency RFID technology, which can improve the security of the access control system, enhance the reliability of the access control system, enhance the intelligence level of the access control system, and improve the management efficiency of the access control system.

[0004] To achieve the above object, the technical solution adopted by the present invention is:

[0005] A door access control method based on ultra-high frequency RFID technology comprises the following steps:

[0006] S1: Set up a multi-antenna array through a distributed layout based on the size, shape and personnel flow of the access control area;

[0007] S2: Based on the Hyper ledger Fabric blockchain platform, nodes are deployed in different locations through a distributed node architecture;

[0008] S3: The artificial intelligence model is trained by collecting data from the access control system. The trained artificial intelligence model analyzes and detects the data in the access control system in real time to identify abnormal behaviors and events.

[0009] S4: Display all data in the access control system in a visual manner on the management platform, and provide data analysis and optimization management based on the authority management mechanism.

[0010] Furthermore, the multi-antenna array is arranged in a distributed manner according to the size, shape and personnel flow of the access control area, including the following steps:

[0011] S11: Determine the antenna of UHF RFID technology based on the gain, directivity and frequency response of the antenna. Set up a multi-antenna array in a distributed layout according to the size, shape and personnel flow of the access control area, and connect the reader and the antenna through a wireless connection;

[0012] S12: preprocessing the signals received by each antenna, measuring and recording the signal strength of the preprocessed signals, obtaining the weight of each antenna through a signal weighting algorithm, and weighted summing the signals;

[0013] S13: Perform independent signal reception, signal comparison and selection, and signal fusion operations through a spatial diversity algorithm, and finally fuse the fused signal obtained by the spatial diversity algorithm and the weighted signal obtained by the signal weighting algorithm through a decision fusion method.

[0014] Furthermore, according to the Hyper ledger Fabric blockchain platform, nodes are deployed in different locations through a distributed node architecture, including the following steps:

[0015] S21: Based on the Hyper ledger Fabric blockchain platform, deploy light nodes on the access control device, deploy full nodes on the central server, and connect the edge smart box between the access control device and the central server;

[0016] S22: Communicate with the UHF RFID reader and light node by configuring the access control management software parameters, manage the full node and communicate with the edge smart box by configuring the blockchain node management software parameters on the central server, and pre-process and encrypt the access control data by configuring the parameters of the data pre-processing and encryption software on the edge smart box;

[0017] S23: When a user approaches the access control device with an access card or tag, the UHF RFID reader reads the user's identity information and sends the information to the access control management software on the access control device. The access control management software sends the user's identity information to the light node. The light node records the information in the local blockchain ledger and sends the information to the full node on the central server. The full node records the received information in the complete blockchain ledger and participates in the consensus process of the blockchain. The edge smart box encrypts and preprocesses the user's identity information, and then sends the encrypted information to the full node on the central server. The full node records the encrypted information in the blockchain ledger.

[0018] S24: After the user's identity information is recorded in the blockchain ledger, the access control management software controls the opening and closing of the access control device according to the user's authority information. If the user has legal authority, the access control device will open the door lock and allow the user to enter; if the user does not have legal authority, the access control device will keep the door lock closed and deny the user entry.

[0019] Furthermore, the full node records the received information in a complete blockchain ledger and participates in the consensus process of the blockchain, including verifying the transaction information received by the full node using a practical Byzantine fault-tolerant consensus algorithm, and coordinating the consensus process through message passing between nodes; the edge smart box pre-processes the user's identity information, including pre-processing the original data from the access control device through data compression and deduplication algorithms; the edge smart box encrypts the user's identity information, including encrypting it through an advanced encryption standard encryption algorithm.

[0020] Furthermore, the artificial intelligence model is trained by collecting data from the access control system, and the trained artificial intelligence model performs real-time analysis and detection on the data in the access control system to identify abnormal behaviors and events, including the following steps:

[0021] S31: pre-processing the collected data according to the data collected from the access control system, wherein the image data is cropped, scaled and grayed, and the time series data is segmented and standardized;

[0022] S32: Divide the preprocessed data into a training set, a validation set, and a test set. Train the artificial intelligence model with the training set. During the training process, optimize the model with the Adam optimization algorithm. Verify the model with the validation set. Evaluate the trained artificial intelligence model with the test set and optimize it based on the evaluation results.

[0023] Furthermore, the artificial intelligence model includes realizing the artificial intelligence model through the fusion of a distributed architecture combined with a convolutional neural network and a recurrent neural network; the distributed architecture includes Apache Dor is.

[0024] Furthermore, the convolutional neural network includes a convolutional neural network model constructed according to the ResNet architecture, a convolutional layer, a pooling layer and a fully connected layer. The convolutional neural network model is trained using preprocessed image data, and the model parameters are adjusted according to a stochastic gradient descent optimization algorithm. The trained convolutional neural network model is applied to new access control system image data to extract the spatial feature representation of the image. The output of the convolutional neural network is a vector feature representing the image.

[0025] Furthermore, the recurrent neural network includes a recurrent neural network model constructed according to a long short-term memory network architecture, a recurrent layer and a fully connected layer. The output of the convolutional neural network is fused with the time series data features through a weighted summation method to obtain an input vector of the recurrent neural network. The recurrent neural network model is trained using the preprocessed time series data and the fused input vector, the model parameters are adjusted according to a back-propagation algorithm, and the spatial features of the convolutional neural network output are combined for analysis and detection. The trained recurrent neural network model is applied to the new access control system time series data and the convolutional neural network output to predict abnormal behaviors and events. The output of the recurrent neural network is a probability value or a classification result.

[0026] Furthermore, the data in the access control system is displayed in a visual manner on the management platform, and data analysis and optimization management are provided according to the authority management mechanism, including the following steps:

[0027] S41: storing data in a manner of combining relational databases and non-relational databases, wherein the relational database is used to store structured data and the non-relational database is used to store unstructured data, and implementing data backup and recovery strategies;

[0028] S42: Display the visual management platform through responsive design. Display the analysis results on the management platform through data visualization technology. The display forms include data charts, maps and reports. Different interface layouts and display contents are displayed according to user needs and authority levels.

[0029] S43: Different permissions are assigned to different user roles through a role-based access control model. Statistical analysis algorithms calculate the average time for people to enter and exit, and indicators of traffic during peak hours. Multi-factor identity verification is performed, including passwords, Bluetooth, facial recognition, and tokens. For abnormal behavior detection results, different colors or icons are used to highlight them.

[0030] S44: Analyze personnel behavior patterns through clustering algorithms and predict future personnel flow through prediction algorithms. Based on the analysis results, discover hidden patterns and trends in the data and generate optimization management recommendation reports.

[0031] An access control system based on ultra-high frequency RFID technology, comprising a multi-antenna array module, a blockchain module, an artificial intelligence algorithm module and a visual management module connected in communication, wherein:

[0032] The multi-antenna array module is used to set up a multi-antenna array through a distributed layout according to the size, shape and personnel flow of the access control area;

[0033] The blockchain module is used to deploy nodes in different locations through a distributed node architecture based on the Hyper ledger Fabric blockchain platform;

[0034] The artificial intelligence algorithm module is used to train the artificial intelligence model by collecting data from the access control system. The trained artificial intelligence model performs real-time analysis and detection on the data in the access control system to identify abnormal behaviors and events.

[0035] The visual management module is used to display the data in the access control system in a visual manner on the management platform, and provide data analysis and optimization management based on the authority management mechanism.

[0036] The beneficial effects of the present invention are:

[0037] 1. By setting up a multi-antenna array through a distributed layout, the signal coverage is expanded, the signal blind area is reduced, and the accuracy and reliability of tag reading are improved. At the same time, the collaborative work of multiple antennas can enhance the signal strength, reduce the risk of interference, and prevent illegal personnel from entering the access control area through interference signals. The authenticity and integrity of the data are ensured through the decentralized, tamper-proof, safe and reliable characteristics of blockchain. Any attempt to tamper with the access control records will be discovered immediately, effectively preventing the data from being maliciously tampered or forged, improving the security of the access control system, and facilitating data exchange with the vehicle management department. By performing real-time analysis and detection of data in the access control system, abnormal behaviors and events can be identified.

[0038] 2. In signal processing and fusion, signal weighting and spatial diversity algorithms are used to improve the quality and stability of the signal. The signal weighting algorithm assigns different weights to different antennas according to the position and signal strength of the antenna, so that antennas with stronger signal strength and higher reliability contribute more to the final fusion result. The spatial diversity algorithm uses the diversity effect of multiple antennas in space to improve the reception quality and reliability of the signal, reduce the probability of signal loss, and improve the reliability of the access control system. On the Hyper ledger Fabric blockchain platform, a distributed node architecture is adopted to deploy nodes in different locations, including light nodes on access control devices and full nodes on central servers. This distributed architecture improves the fault tolerance of the system. Even if a node fails, other nodes can still work normally, ensuring the stable operation of the access control system. The edge smart box connects the access control device and the central server, preprocesses and encrypts the data, and improves the transmission efficiency and security of the data. At the same time, the edge smart box can monitor and analyze the access control data in real time, detect abnormal situations in time and take corresponding measures, further improving the reliability of the access control system.

[0039] 3. The artificial intelligence model obtained by the fusion of two complementary algorithms can give full play to their respective advantages and improve the analysis and detection capabilities of the access control system data. The convolutional neural network and the recurrent neural network are integrated to better analyze and predict personnel behaviors and events. The artificial intelligence algorithm can automatically learn and adapt to different access control scenarios, continuously improve its own performance and accuracy, and enhance the intelligence level of the access control system. The visual management platform displays various data in the access control system in an intuitive way, which is convenient for administrators to manage and monitor. At the same time, the visual management platform provides data analysis and optimization management functions for different users according to the authority management mechanism. Administrators can understand the operating status of the access control system, personnel entry and exit, abnormal behavior detection results, etc. through the visual management platform, make decisions and adjustments in time, and improve the intelligent management level of the access control system. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 The present invention is a flow chart of an access control method based on ultra-high frequency RFID technology.

[0041] Figure 2 It is a schematic diagram of an access control system based on UHF RFID technology. DETAILED DESCRIPTION

[0042] See also Figure 1 and Figure 2 As shown, the present invention relates to an access control method based on ultra-high frequency RFID technology, comprising the following steps:

[0043] S1: Set up a multi-antenna array through a distributed layout based on the size, shape and personnel flow of the access control area;

[0044] S2: Based on the Hyper ledger Fabric blockchain platform, nodes are deployed in different locations through a distributed node architecture;

[0045] S3: The artificial intelligence model is trained by collecting data from the access control system. The trained artificial intelligence model analyzes and detects the data in the access control system in real time to identify abnormal behaviors and events.

[0046] S4: Display all data in the access control system in a visual manner on the management platform, and provide data analysis and optimization management based on the authority management mechanism.

[0047] Furthermore, the multi-antenna array is arranged in a distributed manner according to the size, shape and personnel flow of the access control area, including the following steps:

[0048] S11: Determine the antenna of UHF RFID technology based on the gain, directivity and frequency response of the antenna. Set up a multi-antenna array in a distributed layout according to the size, shape and personnel flow of the access control area, and connect the reader and the antenna through a wireless connection;

[0049] S12: preprocessing the signals received by each antenna, measuring and recording the signal strength of the preprocessed signals, obtaining the weight of each antenna through a signal weighting algorithm, and weighted summing the signals;

[0050] S13: Perform independent signal reception, signal comparison and selection, and signal fusion operations through a spatial diversity algorithm, and finally fuse the fused signal obtained by the spatial diversity algorithm and the weighted signal obtained by the signal weighting algorithm through a decision fusion method.

[0051] Specifically, the advantages of the distributed layout of multiple antenna arrays are that by setting antennas at different locations, all-round coverage of the access control area can be achieved, reducing signal blind spots; multiple antennas working together can increase the signal reception strength and improve the accuracy and reliability of tag reading; multiple tags can be read at the same time, increasing the capacity and processing power of the system; signal interference can be reduced and the system's anti-interference ability can be improved.

[0052] In step S11, according to the layout plan, antennas are installed at different locations in the access control area, ensuring that the distance and angle between antennas are reasonable to avoid signal interference and overlap. According to the reader's reading distance, reading speed and multi-tag processing capability, the UHF RFID reader is determined, and the reader is connected to each antenna to ensure stable and reliable signal transmission. The reader parameters such as frequency, power, sensitivity, etc. are configured to adapt to different application scenarios and environmental conditions.

[0053] In step S12, the signals received by each antenna are preprocessed, including filtering, amplification, denoising and other operations to improve the quality and stability of the signal, and the signal strength of the preprocessed signal is measured and recorded to prepare for the signal weighting algorithm. According to the implementation steps of the signal weighting algorithm, the weight of each antenna is calculated, and the signal is weighted and summed to obtain the weighted signal sum as one of the inputs of signal fusion.

[0054] In step S13, according to the implementation steps of the spatial diversity algorithm, operations such as antenna layout and selection, independent signal reception, signal comparison and selection, and signal fusion are performed. The fused signal obtained by the spatial diversity algorithm and the sum of the weighted signals obtained by the signal weighting algorithm are further fused. The decision fusion is to select the signal obtained by the signal weighting algorithm or the spatial diversity algorithm as the final fused signal according to certain decision rules. For example, if the signal quality obtained by the spatial diversity algorithm is better, then the signal is selected; otherwise, the signal obtained by the signal weighting algorithm is selected. The fused signal is further processed and analyzed to extract the tag information and make a decision on access control.

[0055] Through the combined effect of signal weighting and spatial diversity algorithms, more accurate and reliable signal processing and fusion can be achieved in access control of UHF RFID technology. The signal weighting algorithm can assign different weights according to factors such as antenna signal strength, distance, and reliability, so that antennas with better signal quality contribute more to the final result. The spatial diversity algorithm improves the stability and anti-interference ability of the signal through independent reception and signal comparison and selection of multiple antennas. Combining the two algorithms, we can give full play to their advantages and improve the performance of the access control system.

[0056] Furthermore, according to the Hyper ledger Fabric blockchain platform, nodes are deployed in different locations through a distributed node architecture, including the following steps:

[0057] S21: Based on the Hyper ledger Fabric blockchain platform, deploy light nodes on the access control device, deploy full nodes on the central server, and connect the edge smart box between the access control device and the central server;

[0058] S22: Communicate with the UHF RFID reader and light node by configuring the access control management software parameters, manage the full node and communicate with the edge smart box by configuring the blockchain node management software parameters on the central server, and pre-process and encrypt the access control data by configuring the parameters of the data pre-processing and encryption software on the edge smart box;

[0059] S23: When a user approaches the access control device with an access card or tag, the UHF RFID reader reads the user's identity information and sends the information to the access control management software on the access control device. The access control management software sends the user's identity information to the light node. The light node records the information in the local blockchain ledger and sends the information to the full node on the central server. The full node records the received information in the complete blockchain ledger and participates in the consensus process of the blockchain. The edge smart box encrypts and preprocesses the user's identity information, and then sends the encrypted information to the full node on the central server. The full node records the encrypted information in the blockchain ledger.

[0060] S24: After the user's identity information is recorded in the blockchain ledger, the access control management software controls the opening and closing of the access control device according to the user's authority information. If the user has legal authority, the access control device will open the door lock and allow the user to enter; if the user does not have legal authority, the access control device will keep the door lock closed and deny the user entry.

[0061] Specifically, in step S21, an ultra-high frequency RFID reader and controller are installed on the access control device to ensure that the reader can accurately read the information in the tag and the controller can control the opening and closing of the door lock. A light node is deployed on the access control device, and the parameters of the light node are configured so that it can communicate with the full node on the central server. The light node can be implemented using an embedded device or a small server. The edge smart box is connected between the access control device and the central server, and the parameters of the edge smart box are configured so that it can pre-process and encrypt the data. The edge smart box can be implemented using dedicated hardware devices or software.

[0062] In step S22, configure the parameters of the blockchain network, such as the number of nodes, consensus algorithm, encryption method, etc., and use the official documents and tools of Hyper ledger Fabric to install and configure. Configure the parameters of the access control management software so that the access control device can communicate with the UHF RFID reader and the light node. The access control management software can be implemented using an existing access control management system or custom-developed software. Configure the parameters of the blockchain node management software so that the central server can manage the full node and communicate with the edge smart box. The blockchain node management software can be implemented using the official tools of Hyper ledger Fabric or third-party software. Configure the parameters of the data preprocessing and encryption software so that the edge smart box can preprocess and encrypt the access control data. The data preprocessing and encryption software is implemented through the existing data processing and encryption software.

[0063] In step S24, it also includes: the blockchain node management software on the central server can query and analyze the data in the blockchain account book so that the administrator can manage and monitor the access control. The administrator can query the user's access control records, permission information, etc. through the blockchain node management software, and can also remotely control and manage the access control equipment. The edge smart box monitors and analyzes the access control data in real time to detect abnormal situations in time and take corresponding measures. For example, if the edge smart box detects abnormal access control records or data traffic, it can issue an alarm to notify the administrator to handle it.

[0064] Furthermore, the full node records the received information in a complete blockchain ledger and participates in the consensus process of the blockchain, including verifying the transaction information received by the full node using a practical Byzantine fault-tolerant consensus algorithm, and coordinating the consensus process through message passing between nodes; the edge smart box pre-processes the user's identity information, including pre-processing the original data from the access control device through data compression and deduplication algorithms; the edge smart box encrypts the user's identity information, including encrypting it through an advanced encryption standard encryption algorithm.

[0065] Specifically, the Practical Byzantine Fault Tolerant Consensus Algorithm is a consensus algorithm suitable for distributed systems. It solves the problems of node failure, malicious behavior or erroneous information in distributed systems through the Practical Byzantine Fault Tolerant Consensus Algorithm, and has high security and performance. In access control based on UHF RFID technology, the Practical Byzantine Fault Tolerant Consensus Algorithm ensures the stable operation of the blockchain network and the consistency of data. The Practical Byzantine Fault Tolerant Consensus Algorithm checks whether the transaction format is correct, whether the signature is valid, whether the user identity is legal, etc. Only transactions that pass these verifications will be further processed and participate in the consensus process. And when the full node receives new transaction information, it will broadcast the transaction to other nodes and collect feedback from other nodes. Other nodes will respond to the transaction based on their own ledger status and verification results. If more than two-thirds of the nodes agree to the transaction, it will be considered valid and added to the blockchain ledger. The Practical Byzantine Fault Tolerant Consensus Algorithm ensures the security and reliability of communication between nodes and prevents attacks and interference from malicious nodes. Even if some nodes fail or behave maliciously, the algorithm can ensure the normal operation of the entire system and the consistency of data. In the actual application of access control, various network problems and node failures may occur. The PBFT consensus algorithm has a certain fault tolerance and can maintain the stable operation of the system even when some nodes have problems.

[0066] Data compression and deduplication algorithms can reduce the storage space and transmission time of access control data and improve system performance. In access control based on UHF RFID technology, data compression and deduplication algorithms are used to pre-process the raw data from access control devices. The basic principle of the data compression algorithm is to compress the duplicate data in the access control data to reduce the data storage space; the basic principle of the deduplication algorithm is to remove the duplicate data in the access control data to reduce the data transmission time. In the application, the data compression and deduplication algorithms are optimized and adjusted according to the system needs and performance requirements, such as adjusting the compression ratio, deduplication ratio, algorithm complexity and other parameters to improve the efficiency and performance of the algorithm.

[0067] The Advanced Encryption Standard (AES) encryption algorithm is a widely used symmetric encryption algorithm with high security and performance. In access control based on UHF RFID technology, the AES encryption algorithm can ensure that the encryption and decryption process of access control data is fast and accurate. Its basic principle is: divide the plaintext data into several blocks, each block is 128 bits in size, and use the key to encrypt each block to obtain the ciphertext data. The decryption process is to divide the ciphertext data into several blocks, each block is 128 bits in size, and use the key to decrypt each block to obtain the plaintext data. In the application, the AES encryption algorithm is optimized and adjusted according to the system requirements and performance requirements, such as adjusting the key length, encryption mode, padding method and other parameters to improve the security and performance of the algorithm.

[0068] Furthermore, the artificial intelligence model is trained by collecting data from the access control system, and the trained artificial intelligence model performs real-time analysis and detection on the data in the access control system to identify abnormal behaviors and events, including the following steps:

[0069] S31: pre-processing the collected data according to the data collected from the access control system, wherein the image data is cropped, scaled and grayed, and the time series data is segmented and standardized;

[0070] S32: Divide the preprocessed data into a training set, a validation set, and a test set. Train the artificial intelligence model with the training set. During the training process, optimize the model with the Adam optimization algorithm. Verify the model with the validation set. Evaluate the trained artificial intelligence model with the test set and optimize it based on the evaluation results.

[0071] Specifically, in step S31, various data in the access control system are collected, including but not limited to personnel behavior data (such as passing time, dwell time, movement trajectory, etc.), environmental data (such as temperature, humidity, light, etc.), and equipment status data (such as RFID reader signal strength, access control equipment operation status, etc.). The collected data is preprocessed to improve data quality and availability, including data cleaning (removing noise, outliers, etc.), denoising (using filtering and other technologies), normalization (mapping data to a specific interval), and other operations. For image data, cropping, scaling, grayscale, and other processing can be performed; for time series data, segmentation, standardization, and other operations can be performed.

[0072] In step S2, the model is verified using the validation set. According to the validation results, the structure and parameters of the model are adjusted to improve the performance of the model. The training and validation process is repeated until the model reaches the expected performance indicators. The trained artificial intelligence model is evaluated using the test set, and the model's accuracy, recall rate, F1 value and other indicators are calculated to evaluate the performance and generalization ability of the model. The performance and accuracy of the model are improved by adjusting the model's structure, parameters, algorithms, etc. The trained artificial intelligence model is deployed to the access control system to perform real-time analysis and detection of the data in the access control system. When the model detects abnormal behavior and events, it will issue early warning information in a timely manner to notify the administrator to handle it.

[0073] Furthermore, the artificial intelligence model includes realizing the artificial intelligence model through the fusion of a distributed architecture combined with a convolutional neural network and a recurrent neural network; the distributed architecture includes Apache Dor is.

[0074] Specifically, the distributed architecture is a distributed SQL data warehouse based on the MPP (massively parallel processing) architecture. Optimization is performed based on monitoring results, including adjusting the parallelism, memory allocation, and data partitioning of the job to improve the efficiency and performance of the job.

[0075] Furthermore, the convolutional neural network includes a convolutional neural network model constructed according to the ResNet architecture, a convolutional layer, a pooling layer and a fully connected layer. The convolutional neural network model is trained using preprocessed image data, and the model parameters are adjusted according to a stochastic gradient descent optimization algorithm. The trained convolutional neural network model is applied to new access control system image data to extract the spatial feature representation of the image. The output of the convolutional neural network is a vector feature representing the image.

[0076] Specifically, the parameters of the model are adjusted so that the model can better extract the spatial features of the image data.

[0077] Furthermore, the recurrent neural network includes a recurrent neural network model constructed according to a long short-term memory network architecture, a recurrent layer and a fully connected layer. The output of the convolutional neural network is fused with the time series data features through a weighted summation method to obtain an input vector of the recurrent neural network. The recurrent neural network model is trained using the preprocessed time series data and the fused input vector, the model parameters are adjusted according to a back-propagation algorithm, and the spatial features of the convolutional neural network output are combined for analysis and detection. The trained recurrent neural network model is applied to the new access control system time series data and the convolutional neural network output to predict abnormal behaviors and events. The output of the recurrent neural network is a probability value or a classification result.

[0078] Specifically, the time series data of the access control system is preprocessed, including segmentation, standardization and other operations, to meet the input format of the recurrent neural network. The output of the convolutional neural network is processed as the input of the recurrent neural network. The parameters of the model are adjusted so that the model can better extract the time characteristics of the time series data. In the above model adjustment and optimization, the hyperparameters of the convolutional neural network and the recurrent neural network are adjusted, such as the learning rate, batch size, hidden layer size, etc., to improve the performance and generalization ability of the model; try different fusion methods, such as different splicing methods, weight settings for weighted summation, etc., to find the best fusion method and improve the accuracy and reliability of the model; use the model integration method to combine multiple different convolutional neural network and recurrent neural network models to improve the performance and stability of the model.

[0079] Furthermore, the data in the access control system is displayed in a visual manner on the management platform, and data analysis and optimization management are provided according to the authority management mechanism, including the following steps:

[0080] S41: storing data in a manner of combining relational databases and non-relational databases, wherein the relational database is used to store structured data and the non-relational database is used to store unstructured data, and implementing data backup and recovery strategies;

[0081] S42: Display the visual management platform through responsive design. Display the analysis results on the management platform through data visualization technology. The display forms include data charts, maps and reports. Different interface layouts and display contents are displayed according to user needs and authority levels.

[0082] S43: Different permissions are assigned to different user roles through a role-based access control model. Statistical analysis algorithms calculate the average time for people to enter and exit, and indicators of traffic during peak hours. Multi-factor identity verification is performed, including passwords, Bluetooth, facial recognition, and tokens. For abnormal behavior detection results, different colors or icons are used to highlight them.

[0083] S44: Analyze personnel behavior patterns through clustering algorithms and predict future personnel flow through prediction algorithms. Based on the analysis results, discover hidden patterns and trends in the data and generate optimization management recommendation reports.

[0084] Specifically, in step S41, the relational database is used to store structured data, such as personnel information, access control records, etc.; the non-relational database is used to store unstructured data, such as images, videos, etc. The data backup and recovery strategy can quickly restore data in the event of system failure or data loss to ensure the stable operation of the access control system.

[0085] In step S42, responsive design is a web design and development method, the purpose of which is to enable the web page to automatically adjust the layout and display effect according to different device screen sizes and resolutions to provide a good user experience. Responsive design is adopted to enable the visual management platform to adapt to the screen sizes of different devices, such as computers, tablets, mobile phones, etc. Design a concise and clear interface layout to avoid information overload, and use intuitive charts, graphics and colors to display data.

[0086] In step S43, different permissions are assigned to different user roles. For example, the administrator has the highest permissions and can perform system settings, data management and other operations; ordinary users can only view their own access control records and basic information. User permissions can be managed flexibly to ensure data security and confidentiality. Users with different roles can only access their authorized functions and data, avoiding the risk of misoperation and data leakage. Multi-factor authentication increases the security of the system and prevents illegal users from intruding.

[0087] In step S44, personnel behavior patterns are analyzed through data mining and machine learning algorithms, and access control data is deeply analyzed to provide a variety of visualization chart types, such as bar charts, line charts, pie charts or heat maps, to meet the display requirements of different data. Based on the data analysis results, an optimization management suggestion report is automatically generated, which may include problem analysis, solutions, implementation steps, etc. A feedback mechanism is established to allow users to evaluate and provide feedback on optimization suggestions in order to continuously improve the quality of the suggestions.

[0088] An access control system based on ultra-high frequency RFID technology, comprising a multi-antenna array module, a blockchain module, an artificial intelligence algorithm module and a visual management module connected in communication, wherein:

[0089] The multi-antenna array module is used to set up a multi-antenna array through a distributed layout according to the size, shape and personnel flow of the access control area;

[0090] The blockchain module is used to deploy nodes in different locations through a distributed node architecture based on the Hyper ledger Fabric blockchain platform;

[0091] The artificial intelligence algorithm module is used to train the artificial intelligence model by collecting data from the access control system. The trained artificial intelligence model performs real-time analysis and detection on the data in the access control system to identify abnormal behaviors and events.

[0092] The visual management module is used to display the data in the access control system in a visual manner on the management platform, and provide data analysis and optimization management based on the authority management mechanism.

[0093] The above implementation modes are merely descriptions of the preferred implementation modes of the present invention, and are not intended to limit the scope of the present invention. Without departing from the design spirit of the present invention, various modifications and improvements made to the technical solutions of the present invention by ordinary engineering and technical personnel in the field shall fall within the protection scope determined by the claims of the present invention.

Claims

1. A door access control method based on ultra-high frequency RFID technology, characterized in that: The following steps are involved: S1: Set up a multi-antenna array through a distributed layout based on the size, shape and personnel flow of the access control area; S2: Based on the Hyperledger Fabric blockchain platform, nodes are deployed in different locations through a distributed node architecture; S3: The artificial intelligence model is trained by collecting data from the access control system. The trained artificial intelligence model analyzes and detects the data in the access control system in real time to identify abnormal behaviors and events. S4: Display all data in the access control system in a visual manner on the management platform, and provide data analysis and optimization management based on the authority management mechanism.

2. The access control method based on ultra-high frequency RFID technology according to claim 1 is characterized in that: The method of setting up a multi-antenna array through a distributed layout according to the size, shape and personnel flow of the access control area includes the following steps: S11: Determine the antenna of UHF RFID technology based on the gain, directivity and frequency response of the antenna. Set up a multi-antenna array in a distributed layout according to the size, shape and personnel flow of the access control area, and connect the reader and the antenna through a wireless connection; S12: preprocessing the signals received by each antenna, measuring and recording the signal strength of the preprocessed signals, obtaining the weight of each antenna through a signal weighting algorithm, and weighted summing the signals; S13: Perform independent signal reception, signal comparison and selection, and signal fusion operations through a spatial diversity algorithm, and finally fuse the fused signal obtained by the spatial diversity algorithm and the weighted signal obtained by the signal weighting algorithm through a decision fusion method.

3. The access control method based on ultra-high frequency RFID technology according to claim 1 is characterized in that: According to the Hyperledger Fabric blockchain platform, nodes are deployed in different locations through a distributed node architecture, including the following steps: S21: Based on the Hyperledger Fabric blockchain platform, deploy light nodes on the access control device, deploy full nodes on the central server, and connect the edge smart box between the access control device and the central server; S22: Communicate with the UHF RFID reader and light node by configuring the access control management software parameters, manage the full node and communicate with the edge smart box by configuring the blockchain node management software parameters on the central server, and pre-process and encrypt the access control data by configuring the parameters of the data pre-processing and encryption software on the edge smart box; S23: When a user approaches the access control device with an access card or tag, the UHF RFID reader reads the user's identity information and sends the information to the access control management software on the access control device. The access control management software sends the user's identity information to the light node. The light node records the information in the local blockchain ledger and sends the information to the full node on the central server. The full node records the received information in the complete blockchain ledger and participates in the consensus process of the blockchain. The edge smart box encrypts and preprocesses the user's identity information, and then sends the encrypted information to the full node on the central server. The full node records the encrypted information in the blockchain ledger. S24: After the user's identity information is recorded in the blockchain ledger, the access control management software controls the opening and closing of the access control device according to the user's authority information. If the user has legal authority, the access control device will open the door lock and allow the user to enter; if the user does not have legal authority, the access control device will keep the door lock closed and deny the user entry.

4. The access control method based on UHF RFID technology according to claim 3 is characterized in that: The full node records the received information in the complete blockchain ledger and participates in the consensus process of the blockchain, including verifying the transaction information received by the full node using the Byzantine fault-tolerant consensus algorithm and coordinating the consensus process through message passing between nodes; the edge smart box pre-processes the user's identity information, including pre-processing the original data from the access control device through data compression and deduplication algorithms; The edge smart box encrypts the user's identity information, including encryption through the Advanced Encryption Standard encryption algorithm.

5. The access control method based on UHF RFID technology according to claim 1 is characterized in that: The artificial intelligence model is trained by collecting data from the access control system, and the trained artificial intelligence model performs real-time analysis and detection on the data in the access control system to identify abnormal behaviors and events, including the following steps: S31: pre-processing the collected data according to the data collected from the access control system, wherein the image data is cropped, scaled and grayed, and the time series data is segmented and standardized; S32: Divide the preprocessed data into a training set, a validation set, and a test set. Train the artificial intelligence model with the training set. During the training process, optimize the model with the Adam optimization algorithm. Verify the model with the validation set. Evaluate the trained artificial intelligence model with the test set and optimize it based on the evaluation results.

6. The access control method based on UHF RFID technology according to claim 5 is characterized in that: The artificial intelligence model includes realizing the artificial intelligence model by combining a distributed architecture with a convolutional neural network and a recurrent neural network; the distributed architecture includes Apache Doris.

7. The access control method based on UHF RFID technology according to claim 6 is characterized in that: The convolutional neural network includes a convolutional neural network model constructed according to the ResNet architecture, a convolutional layer pooling layer and a fully connected layer. The convolutional neural network model is trained by preprocessed image data, and the model parameters are adjusted according to a stochastic gradient descent optimization algorithm. The trained convolutional neural network model is applied to new access control system image data to extract the spatial feature representation of the image. The output of the convolutional neural network is a vector feature representing the image.

8. The access control method based on UHF RFID technology according to claim 6 is characterized in that: The recurrent neural network includes a recurrent neural network model constructed according to a long short-term memory network architecture, a recurrent layer and a fully connected layer. The output of the convolutional neural network is fused with the time series data features through a weighted summation method to obtain an input vector of the recurrent neural network. The recurrent neural network model is trained through the preprocessed time series data and the fused input vector. The model parameters are adjusted according to a back propagation algorithm. The spatial features of the convolutional neural network output are combined for analysis and detection. The trained recurrent neural network model is applied to the new access control system time series data and the convolutional neural network output to predict abnormal behaviors and events. The output of the recurrent neural network is a probability value or a classification result.

9. The access control method based on ultra-high frequency RFID technology according to claim 1, characterized in that: The method of displaying the data in the access control system in a visual manner on the management platform and providing data analysis and optimization management according to the authority management mechanism includes the following steps: S41: storing data in a manner of combining relational databases and non-relational databases, wherein the relational database is used to store structured data and the non-relational database is used to store unstructured data, and implementing data backup and recovery strategies; S42: Display the visual management platform through responsive design. Display the analysis results on the management platform through data visualization technology. The display forms include data charts, maps and reports. Different interface layouts and display contents are displayed according to user needs and authority levels. S43: Different permissions are assigned to different user roles through a role-based access control model. Statistical analysis algorithms calculate the average time for people to enter and exit, and indicators of traffic during peak hours. Multi-factor identity verification is performed, including passwords, Bluetooth, facial recognition, and tokens. For abnormal behavior detection results, different colors or icons are used to highlight them. S44: Analyze personnel behavior patterns through clustering algorithms and predict future personnel flow through prediction algorithms. Based on the analysis results, discover hidden patterns and trends in the data and generate optimization management recommendation reports.

10. An access control system based on ultra-high frequency RFID technology, used to execute an access control method based on ultra-high frequency RFID technology as claimed in any one of claims 1 to 9, characterized in that: It includes a multi-antenna array module with communication connection, a blockchain module, an artificial intelligence algorithm module and a visual management module, among which: The multi-antenna array module is used to set up a multi-antenna array through a distributed layout according to the size, shape and personnel flow of the access control area; The blockchain module is used to deploy nodes in different locations through a distributed node architecture based on the Hyperledger Fabric blockchain platform; The artificial intelligence algorithm module is used to train the artificial intelligence model by collecting data from the access control system. The trained artificial intelligence model performs real-time analysis and detection on the data in the access control system to identify abnormal behaviors and events. The visual management module is used to display the data in the access control system in a visual manner on the management platform, and provide data analysis and optimization management based on the authority management mechanism.

Images