Method and system for identifying tag in RFID system

Abstract

The invention discloses a method and system for identifying tags in an RFID system, and relates to the field of tag identification, and the method comprises the following steps: 1, collecting RFID tag communication data, extracting signal modes of receiving and transmitting links in normal operation, constructing a deep learning model, and identifying key features of the tag communication data in the corresponding signal modes, taking the key features as training samples of a deep learning model; 2, running a deep learning model, traversing a communication link, and extracting key features in the communication link; step 3, continuously monitoring real-time data of a communication link and capturing abnormal fluctuation in a signal; key features are automatically identified and data classification is performed through a deep learning model, deeper features are learned and extracted from complex data, the intelligent level of the system is improved, and a model with high generalization ability is established through specially designed feature engineering and proper training sample division.

Description

Technical Field

[0001] The present invention relates to the technical field of tag identification, and in particular to a method and system for identifying tags in an RFID system. Background Art

[0002] RFID system generally consists of three parts: signal transmitter, signal receiver, and transmitting and receiving antenna. In RFID system, signal transmitter exists in different forms for different application purposes. The typical form is tag. Tag is equivalent to barcode symbol in barcode technology, which is used to store information that needs to be identified and transmitted. Signal receiver is generally called reader. The basic function of reader is to provide a way to transmit data with tag. In addition, reader also provides quite complex signal status control, parity error checking and correction functions, etc. In addition to storing the information that needs to be transmitted, the tag must also contain certain additional information, such as error checking information, etc. However, the existing RIFD system still has some shortcomings during use, such as: 1. Traditional RFID systems lack sufficient intelligence and cannot effectively and accurately identify anomalies from massive data. Anomaly detection relies on simple rules or thresholds, which are prone to false positives or false negatives, especially in dynamically changing environments. 2. Traditional processing methods are insufficient in dealing with abnormal situations such as complex signal interference and data tampering. Inefficient data processing and analysis methods will lead to high latency and response time, affecting system performance. Summary of the invention

[0003] 1. Technical issues to be resolved In view of the above-mentioned shortcomings of the prior art, the present invention provides a method and system for identifying tags in an RFID system, which can effectively solve the problems of the prior art.

[0004] (II) Technical solution To achieve the above objectives, the present invention is implemented through the following technical solutions: The present invention discloses a method for identifying a tag in an RFID system, comprising the following steps: Step 1: Collect RFID tag communication data, extract the signal patterns of the receiving and transmitting links in normal operation, build a deep learning model, and identify the key features of the tag communication data under the corresponding signal pattern, and use the key features as training samples for the deep learning model; Step 2: Run the deep learning model, traverse the communication links, and extract key features in the communication links; Step 3: Continuously monitor the real-time data of the communication link and capture abnormal fluctuations in the signal; Step 4: When an abnormal fluctuation is captured, the label data related to the abnormal fluctuation is parsed and compared with the pre-set standard format to determine whether there is an error; Step 5: If the label data related to the analyzed abnormal fluctuations show an error that is inconsistent with the preset standard, trace the potential cause of the error; Step 6: Issue warnings for potential causes and generate corresponding corrective plans. Build a prediction model, use the key characteristic parameters of the corrective plan as samples, simulate the corrective plan, evaluate the corrective effect of the corrective plan, and rank and feedback the corrective effects.

[0005] Furthermore, the attributes of the key features in step 1 include: signal strength, modulation and demodulation characteristics, transmission time and error code.

[0006] Furthermore, the abnormal fluctuation attributes in step 3 include: signal loss, abnormal signal fluctuation, data packet tampering and data packet damage.

[0007] Furthermore, the potential causes in step 5 include: antenna damage, tag damage, firmware abnormality, data processing algorithm abnormality, signal source electromagnetic interference, signal attenuation caused by changes in humidity, temperature and physical obstacles.

[0008] Furthermore, the prediction model in step 6 automatically labels the predicted data as normal and abnormal situations, and when the abnormal situation is labeled, it is submitted for iterative update identification of the signal pattern.

[0009] The identification tag system in the RFID system includes: Data acquisition module, used to collect communication data of RFID tags under normal and abnormal conditions, and extract signal strength, modulation and demodulation characteristics, transmission time and error code; A data processing module is used to pre-process the collected data and divide the collected data into a training set and a test set; The training module is used to cyclically train the deep learning model using the training set, optimize it according to the preset training cycle, and consider the model to be usable after the performance evaluation of the model meets the preset threshold; The capture module is used to monitor the communication link data in real time, extract key features in the communication link through the model, and capture abnormal fluctuations; The judgment module is used to analyze the label content corresponding to the abnormal data captured, compare it with the pre-set standard format, and determine whether there is an error; The traceability module is used to initiate traceability analysis to identify the cause of the error when there is an error in the data, and to analyze hardware failures, software anomalies, and environmental change parameters.

[0010] Furthermore, the traceability module is interactively connected to a correction module via a wireless network. The correction module issues warnings for detected problems, generates corresponding correction strategies, develops prediction models, automatically monitors and annotates data, and the annotated results are used for decision support.

[0011] Furthermore, the correction module is interactively connected to an evaluation module via a wireless network, and the evaluation module is used to submit iterative updates and correction plans for the corresponding signal patterns, simulate new feature parameters as samples, evaluate and provide feedback on the operating effects of the correction plans, and use the feedback parameters as reference values ​​for sorting.

[0012] Furthermore, the data acquisition module and the data processing module are interactively connected through a wireless network, the data processing module and the training module are interactively connected through a wireless network, the training module and the capture module are interactively connected through a wireless network, the capture module and the judgment module are interactively connected through a wireless network, and the judgment module and the traceability module are interactively connected through a wireless network.

[0013] (III) Beneficial effects Compared with the prior art, the technical solution provided by the present invention has the following beneficial effects: Through the deep learning model, key features are automatically identified and data is classified, and deeper features are learned and extracted from complex data, which improves the intelligence level of the system. Through specially designed feature engineering and appropriate training sample division, a model with high generalization ability is established.

[0014] By continuously monitoring real-time data and quickly capturing and responding to abnormal signals, the ability to respond to abnormal label identification is significantly improved. The constructed predictive model can automatically identify normal and abnormal label data, providing a mechanism to identify and analyze the root causes of abnormal data and create targeted corrective solutions, which not only solves immediate problems but also helps prevent the same problems in the future.

[0015] Through traceability analysis, adaptive adjustments are made to various causes, which enhances the system's fault tolerance, improves the system's stability and security, and also enables the RFID system to automatically adapt to complex and changing application scenarios. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the prior art descriptions. Obviously, the drawings described below are only some embodiments of the present invention, and for ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0017] Figure 1 A schematic diagram of a flow chart of a method for identifying a tag in an RFID system in the present invention; Figure 2 It is a schematic diagram of the framework of the identification tag system in the RFID system of the present invention; The numbers in the figure represent: 1. Data acquisition module; 2. Data processing module; 3. Training module; 4. Capture module; 5. Judgment module; 6. Tracing module; 7. Correction module; 8. Evaluation module. DETAILED DESCRIPTION

[0018] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0019] The present invention will be further described below in conjunction with the embodiments.

[0020] The method for identifying a tag in an RFID system of this embodiment is as follows: Figure 1 As shown, the following steps are included: Step 1: Collect RFID tag communication data, extract the signal patterns of the receiving and transmitting links in normal operation, build a deep learning model, and identify the key features of the tag communication data under the corresponding signal pattern. The attributes of the key features include: signal strength, modulation and demodulation characteristics, transmission time and error code. The key features are used as training samples for the deep learning model; the training samples are divided into a test set and a training set. The training set is cyclically put into the deep learning model construction, and after the preset training cycle is completed, it is put into the test set for testing. The test results are evaluated. After the evaluation participation reaches or exceeds the preset threshold, the deep learning model is considered to be in an available state; Step 2: Run the deep learning model, traverse the communication links, and extract key features in the communication links; Step 3: Continuously monitor the real-time data of the communication link and capture abnormal fluctuations in the signal. The abnormal fluctuation attributes include: signal loss, abnormal signal fluctuation, data packet tampering and data packet damage; Step 4: When an abnormal fluctuation is captured, the label data related to the abnormal fluctuation is parsed and compared with the pre-set standard format to determine whether there is an error; Step 5: If the tag data related to the analyzed abnormal fluctuations show an error that is inconsistent with the preset standard, trace the source and analyze the potential causes of the error, which may include: antenna damage, tag damage, firmware abnormality, data processing algorithm abnormality, signal source electromagnetic interference, humidity, temperature, and signal attenuation caused by changes in physical obstacles; Step 6: Issue warnings for potential causes and generate corresponding corrective plans. Build a prediction model that automatically labels normal and abnormal situations for the predicted data. When the abnormal situation is labeled, submit the iterative update identification of the signal pattern, use the key characteristic parameters of the corrective plan as samples, simulate the corrective plan, evaluate the corrective effect of the corrective plan, and rank the corrective effect for feedback.

[0021] Compared with existing technologies, the introduction of deep learning models makes it possible to extract key features from complex communication data, which can more accurately identify RFID tags. Even in environments with complex signal patterns or noise, through real-time monitoring and automatic capture of abnormal fluctuations, the system can quickly identify potential problems such as signal loss, abnormal signal fluctuations, etc., so as to respond immediately and ensure system stability and data integrity; Through continuous training and testing, as well as iterative updates to identify abnormal fluctuations, the model can adaptively optimize its performance and maintain the ability to identify new and complex attacks. It provides a mechanism to identify and analyze the root causes of abnormal data and create targeted corrective solutions, which not only solves immediate problems but also helps prevent the same problems in the future. The use of predictive models to automatically annotate data, combined with the high accuracy of deep learning models, reduces false positives and missed negatives and improves the overall recognition accuracy of the system. Corrective actions will be simulated and evaluated to ensure that the measures taken are effective. It also provides a feedback link to continuously optimize and improve system performance.

[0022] In other aspects, this embodiment also provides an identification tag system in an RFID system, such as Figure 2 As shown, including: Data acquisition module 1, used to collect communication data of RFID tags under normal and abnormal conditions, and extract signal strength, modulation and demodulation characteristics, transmission time and error code; Data processing module 2, used to pre-process the collected data and divide the collected data into a training set and a test set; Training module 3, used to cyclically train the deep learning model using the training set, optimize according to a preset training cycle, and consider the model to be in a usable state after the model meets the performance evaluation of a preset threshold; Capture module 4, used to monitor communication link data in real time, extract key features in the communication link through the model, and capture abnormal fluctuations; The judgment module 5 is used to analyze the label content corresponding to the abnormal data captured, compare it with the pre-set standard format, and judge whether there is an error; The traceability module 6 is used to initiate traceability analysis to identify the cause of the error when there is an error in the data, and to analyze hardware failures, software anomalies and environmental change parameters.

[0023] As a preferred implementation in this embodiment, Figure 2 As shown, the data acquisition module 1 and the data processing module 2 are interactively connected through a wireless network, the data processing module 2 and the training module 3 are interactively connected through a wireless network, the training module 3 and the capture module 4 are interactively connected through a wireless network, the capture module 4 and the judgment module 5 are interactively connected through a wireless network, and the judgment module 5 and the traceability module 6 are interactively connected through a wireless network.

[0024] In this embodiment, by introducing deep learning and real-time monitoring technology, the intelligence and adaptability of the RFID system are improved, the detection, analysis and processing capabilities of various abnormal conditions are enhanced, and the efficiency and reliability of the RFID system are significantly improved. It has important application value in modern supply chain management, inventory tracking, anti-counterfeiting and security.

[0025] In this embodiment, Figure 2 As shown, the traceability module 6 is interactively connected to the correction module 7 via a wireless network. The correction module 7 issues warnings for detected problems, generates corresponding correction strategies, formulates prediction models, automatically monitors and annotates data, and the annotation results are used for decision support. The correction module 7 is interactively connected to the evaluation module 8 via a wireless network. The evaluation module 8 is used to submit iterative updates and correction plans for corresponding signal patterns, simulate new feature parameters as samples, evaluate and provide feedback on the operating effects of the correction plans, and use the feedback parameters as reference values ​​for sorting.

[0026] Compared with existing technologies, it can automatically monitor and annotate data for decision support. This intelligent function plays a key role in monitoring anomalies and automated processing. After detecting a problem, the system can not only issue a warning, but also generate a correction strategy and implement it quickly. It further uses iterative updates to ensure that the system is constantly optimized and adapts to new situations. Through the evaluation and feedback of the strategy operation effect, the system can self-optimize, and dynamically adjust and continuously improve performance based on feedback parameters.

[0027] In summary, the present invention uses a deep learning model to automatically identify key features and classify data, deeply explore deep-level features in complex data, and thus improve the intelligence of the system. Through carefully designed feature engineering and reasonable division of training samples, a model with high generalization ability is constructed; At the same time, the present invention also realizes continuous monitoring of real-time data, can quickly capture and respond to abnormal signals, and significantly improves the recognition and response speed of abnormal labels. The established prediction model can not only automatically distinguish between normal and abnormal label data, but also provide a mechanism for identifying and analyzing the root causes of abnormal data, and formulate targeted corrective measures accordingly, which not only effectively solves the immediate problems, but also provides strong support for preventing similar problems in the future; In addition, through traceability analysis, the present invention can make adaptive adjustments according to different causes, enhance the fault tolerance of the system, and improve the stability and security of the system. This enables the RFID system to better adapt to complex and changeable application scenarios and provides more reliable technical support for practical applications.

[0028] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, a person skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements will not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for identifying a tag in an RFID system, characterized in that: The following steps are involved: Step 1: Collect RFID tag communication data, extract the signal patterns of the receiving and transmitting links in normal operation, build a deep learning model, and identify the key features of the tag communication data under the corresponding signal pattern, and use the key features as training samples for the deep learning model; Step 2: Run the deep learning model, traverse the communication links, and extract key features in the communication links; Step 3: Continuously monitor the real-time data of the communication link and capture abnormal fluctuations in the signal; Step 4: When an abnormal fluctuation is captured, the label data related to the abnormal fluctuation is parsed and compared with the pre-set standard format to determine whether there is an error; Step 5: If the label data related to the analyzed abnormal fluctuations show an error that is inconsistent with the preset standard, trace the potential cause of the error; Step 6: Issue warnings for potential causes and generate corresponding corrective plans. Build a prediction model, use the key characteristic parameters of the corrective plan as samples, simulate the corrective plan, evaluate the corrective effect of the corrective plan, and rank and feedback the corrective effects.

2. The method for identifying a tag in an RFID system according to claim 1, characterized in that: In step 1, the training samples are divided into a test set and a training set. The training set is cyclically put into the deep learning model construction, and after the preset training cycle is completed, it is put into the test set for testing. The test results are evaluated, and after the evaluation participation reaches or exceeds the preset threshold, the deep learning model is considered to be in an available state.

3. The method for identifying a tag in an RFID system according to claim 1, characterized in that: The attributes of the key features in step 1 include: signal strength, modulation and demodulation characteristics, transmission time and error code.

4. The method for identifying a tag in an RFID system according to claim 1, characterized in that: The abnormal fluctuation attributes in step 3 include: signal loss, abnormal signal fluctuation, data packet tampering and data packet damage.

5. The method for identifying a tag in an RFID system according to claim 1, characterized in that: Potential causes in step 5 include: antenna damage, tag damage, firmware abnormality, data processing algorithm abnormality, signal source electromagnetic interference, and signal attenuation caused by changes in humidity, temperature, and physical obstacles.

6. The method for identifying a tag in an RFID system according to claim 1, characterized in that: The prediction model in step 6 automatically labels the predicted data as normal and abnormal situations, and when the abnormal situation is labeled, it is submitted for iterative update identification of the signal pattern.

7. An identification tag system in an RFID system, the system being an implementation system of the method for identifying tags in an RFID system according to any one of claims 1 to 6, characterized in that: include: A data acquisition module (1) is used to collect communication data of RFID tags under normal and abnormal conditions, and extract signal strength, modulation and demodulation characteristics, transmission time and error code; A data processing module (2) is used to pre-process the collected data and divide the collected data into a training set and a test set; A training module (3) is used to cyclically train the deep learning model using the training set, optimize the model according to a preset training cycle, and consider the model to be in a usable state after the model meets the performance evaluation of a preset threshold; A capture module (4) is used to monitor the communication link data in real time, extract key features in the communication link through the model, and capture abnormal fluctuations; A judgment module (5) is used to analyze the label content corresponding to the abnormal data of the captured abnormal fluctuation, compare it with a pre-set standard format, and judge whether there is an error; The traceability module (6) is used to initiate traceability analysis to identify the cause of the error when there is an error in the data, and to analyze hardware failures, software anomalies and environmental change parameters.

8. The identification tag system in the RFID system according to claim 7, characterized in that: The traceability module (6) is interactively connected to the correction module (7) via a wireless network. The correction module (7) issues a warning for the detected problem, generates a corresponding correction strategy, formulates a prediction model, and automatically monitors and annotates the data. The annotated results are used for decision support.

9. The identification tag system in the RFID system according to claim 8, characterized in that: The correction module (7) is interactively connected to an evaluation module (8) via a wireless network. The evaluation module (8) is used to submit iterative updates and correction plans for corresponding signal patterns, simulate new characteristic parameters as samples, evaluate and provide feedback on the running effects of the correction plans, and use the feedback parameters as reference values ​​for sorting.

10. The identification tag system in the RFID system according to claim 7, characterized in that: The data acquisition module (1) is interactively connected to the data processing module (2) via a wireless network, the data processing module (2) is interactively connected to the training module (3) via a wireless network, the training module (3) is interactively connected to the capture module (4) via a wireless network, the capture module (4) is interactively connected to the judgment module (5) via a wireless network, and the judgment module (5) is interactively connected to the tracing module (6) via a wireless network.

Images