Double-RFID liquid leakage detection method and device based on multi-frequency feature fusion

Abstract

The invention provides a double-RFID liquid leakage detection method and device based on multi-frequency feature fusion, and the method comprises the steps: enabling a pair of ultrahigh-frequency RFID tags to be parallelly, closely and fixedly disposed as a double-tag sensor, arranging a liquid absorption medium between the two tags, capturing leaked liquid, and amplifying the influence; constructing a liquid leakage sensing model based on the mutual inductance coupling effect between the antenna circuits adjacent to the tags; the frequency hopping characteristic in the EPC Gen2 protocol is utilized to uniformly analyze the phases of a plurality of label signals under multiple frequencies, and the characteristic with strong robustness is extracted; and fusing a plurality of extracted features, carrying out comprehensive analysis on feature vectors by adopting a decision tree classification model, and distinguishing real leakage, environmental interference and sensor static change. According to the method, the medium change caused by liquid leakage is sensed by utilizing the coupling characteristic of the adjacent RFID tags, the multi-frequency phase data is acquired by combining an RFID frequency hopping mechanism, and the high-precision detection of the liquid leakage state is realized by extracting the characteristics through linear fitting.

Description

Technical Field

[0001] This invention relates to the field of industrial testing technology, and in particular, to a leakage detection technology based on dual RFID tags and multi-frequency phase fitting; more specifically, it relates to a dual RFID leakage detection method and apparatus based on multi-frequency feature fusion. Background Technology

[0002] Liquid leakage is a common safety hazard and source of economic loss in modern industrial equipment, pipelines, and liquid storage devices. For example, cooling systems, lubricating oil pipelines, and chemical storage tanks are prone to leakage during long-term operation due to aging seals or loose connections. Once a leak occurs, it not only wastes energy and raw materials but may also trigger serious accidents such as fires and explosions. Therefore, how to achieve low-cost, real-time, and stable leak detection has always been an important research direction in industrial safety monitoring.

[0003] Existing leak detection methods mainly include solutions based on vision, chemical sensors, or dedicated pressure / flow sensors, as well as manual inspection. However, vision-based methods are difficult to operate in low-light or non-line-of-sight environments, chemical and flow sensors are typically complex to deploy, costly, and inconvenient to maintain, while manual inspection suffers from significant delays and poor accuracy.

[0004] Radio Frequency Identification (RFID) technology is widely used for object identification and environmental sensing due to its passive nature, low cost, and ease of deployment. Research shows that electromagnetic coupling exists between adjacent RFID tags, and changes in the surrounding medium (such as changes in dielectric constant due to liquid seepage) can cause significant changes in coupling characteristics. However, most existing RFID methods rely solely on Received Signal Strength Indication (RSSI) or single-frequency phase characteristics as detection standards, making them susceptible to environmental noise and random interference, resulting in insufficient detection accuracy and robustness. Summary of the Invention

[0005] Therefore, the purpose of this invention is to develop and design a dual RFID leak detection method and device based on multi-frequency feature fusion. This method does not rely on the absolute value of a single tag signal. Based on the coupling effect of dual RFID tags, it utilizes the coupling characteristics of adjacent RFID tags to sense changes in the medium caused by liquid leakage, monitors the relative relationship between the signals of two adjacent tags, effectively distinguishes leak events from environmental interference, and significantly reduces the false alarm rate. Furthermore, it combines RFID frequency hopping mechanism to obtain multi-frequency phase data, extracts features through linear fitting, and achieves high-precision detection of leak conditions. Moreover, it requires no additional hardware, has high sensitivity, strong real-time performance, and good applicability.

[0006] This invention provides a dual RFID leakage detection method based on multi-frequency feature fusion, comprising the following steps:

[0007] S1. A pair of ultra-high frequency (UHF) RFID tags are fixedly placed in parallel and close proximity as a dual-tag sensor, and a liquid absorption medium is set between the two RFID tags to capture leaked liquid and amplify its effects.

[0008] Specifically, UHF RFID systems have a built-in frequency hopping mechanism, allowing readers to switch between multiple channels. This enables the acquisition of phase information at multiple frequencies without the need for additional hardware, providing a basis for constructing detection methods based on multi-frequency phase fitting. By fitting the linear relationship between phase and frequency at multiple frequencies, a slope factor and goodness of fit reflecting the electromagnetic environment characteristics of the tag can be obtained, thus more accurately characterizing the medium disturbance caused by liquid leakage.

[0009] S2. Based on the mutual inductance coupling effect between the antenna circuits of adjacent RFID tags, a leakage sensing model between the two RFID tags is constructed. The leakage sensing model is characterized as a mathematical model that causes a change in the signal phase of the RFID tag itself and the adjacent RFID tags when the impedance of one RFID tag changes.

[0010] Specifically, when two RFID tags are placed close together in physical space, their antenna circuits are no longer independent of each other, but form a coupled system through electromagnetic fields. When the impedance of one RFID tag changes, it causes a phase change in the signal of itself and the adjacent RFID tags.

[0011] S3. Utilizing the frequency hopping feature in the EPC Gen2 protocol, perform unified analysis on the signal phases of multiple RFID tags at multiple frequencies to extract robust features, including: based on signal phase... With operating frequency A linear relationship exists: , where the slope represents the rate of phase change with frequency, affected by communication distance and antenna equivalent input impedance; b is the intercept of the phase-frequency curve; multiple data points collected within one frequency hopping cycle... Linear fitting is performed on the data points to obtain the key feature parameter, slope. and goodness of fit According to the slope and / or goodness of fit Determine whether the dielectric properties of the dual-tag sensor have been altered due to liquid leakage;

[0012] S4. The extracted features are fused together, and a decision tree classification model is used to comprehensively analyze the feature vectors to effectively distinguish between real leakage, environmental interference, and static changes of the sensor.

[0013] Further, step S2 includes:

[0014] The interaction between two adjacent RFID tags, A and B, can be simplified into an equivalent circuit model. The current I flowing through tag A and tag B is... A and I B The relational equation followed is: ,

[0015] (1)

[0016] In equation (1), , These are the induced source voltages on label A and label B, respectively;

[0017] , The chip impedances are for labels A and B, respectively.

[0018] , These are the impedances of the antennas of tag A and tag B, respectively.

[0019] , These are the mutual impedances between label A and label B, respectively.

[0020] Solving equation (1), we obtain the current I of tag A and tag B. A and I B ;

[0021] The current of any tag is determined by all impedance parameters in the system. Changing any impedance value (such as...) This will simultaneously affect and , and The relationship with impedance is:

[0022] ,

[0023] (2)

[0024] In equation (2), and These represent the total impedances of label A and label B, respectively. and The calculation expressions are as follows:

[0025] ,

[0026] (3)

[0027] The signal phase of an RFID tag is determined by the complex amplitude of the antenna current. The current phase values ​​for tag A and tag B are as follows:

[0028] ,

[0029] (4)

[0030] When an external disturbance is applied to one of the tags, causing changes in the tag's own impedance and mutual impedance, the current phase of both tags will undergo predictable changes.

[0031] For example, when an external disturbance is applied to tag B, it causes its own impedance to be affected. and mutual impedance When this changes, it will cause the current phase of the two tags to change. and All of these undergo predictable changes, allowing physical events to be inferred by measuring the phase matrix;

[0032] Furthermore, the method for determining whether the dielectric properties of the dual-tag sensor have changed due to liquid leakage based on the slope k in step S3 includes:

[0033] The rate of change of phase with frequency (linear rate of change) k of the phase-frequency curve is used as an indicator of leakage status, and the slope is... ,in, It is a component determined by the signal propagation distance. It is a component determined by the RFID tag hardware (especially the antenna equivalent input impedance);

[0034] In the dual-tag sensor of this invention, two tags are fixedly placed, therefore their The components are stable and nearly equal. Errors introduced by the reader hardware. It is also common mode. Therefore,

[0035] Two RFID tags were fixed in place. The components are stable and equal. Component common mode; slope values ​​for two RFID tags. and Perform a difference operation. and The difference between them is slope difference The difference in impedance components between two RFID tags The decision was made to place the product in a dry state. It is a stable benchmark value, thus eliminating the influence of propagation distance and reader hardware;

[0036] When a liquid wets a dual-tag sensor, it asymmetrically changes the dielectric constant around the two tag antennas, causing different changes in the equivalent input impedance of the two tag antennas, resulting in... and The changes are no longer synchronized, causing the slope difference to... Deviation from its baseline value in the dry state; by monitoring the slope difference between the two labels. Whether there is a significant deviation from a stable reference value can uniquely and reliably determine whether the dielectric properties of the dual-tag sensor have changed due to liquid leakage.

[0037] Furthermore, the S3 step is based on the goodness of fit. Methods for determining whether liquid leakage has altered the dielectric properties of a dual-tag sensor include:

[0038] Based on goodness of fit (Coefficient of Determination) reflects the degree of fit between the data points and the fitted linear model, and is expressed as goodness of fit. As a marker to determine whether a leakage process is occurring dynamically;

[0039] The goodness of fit The calculation methods include: obtaining a straight line by fitting using the least squares method. ;in It is the model prediction value of the signal phase. It is the actual frequency;

[0040] Calculate observation data Relative to its average value Sum of squares of total variation:

[0041] ;

[0042] Calculate the actual observed values and model predictions The sum of squares (SSE) of the differences (i.e., residuals) between them;

[0043] ;

[0044] The goodness of fit is calculated based on SST and SSE. :

[0045] ;

[0046] like If the value drops significantly, it is determined that a leakage process is occurring dynamically.

[0047] Specifically, under stable conditions of dryness or liquid saturation, the impedance of the RFID tag is constant, and the data collected within a complete frequency hopping cycle... The data points can well conform to a single linear relationship, and the calculated values ​​are... The value is close to 1.

[0048] Liquid leakage is a dynamic process, with the amount of liquid continuously increasing. This means that the tag's impedance changes continuously during the brief time interval of frequency hopping measurement. Therefore, the phase measured at different frequency points actually corresponds to the state at different times and under different impedances.

[0049] When linearly fitting these data points, which correspond to dynamically changing states, they will no longer lie strictly on a straight line, causing the data points to exhibit a discrete or non-linear trend. This will worsen the linear fitting effect, resulting in unreliable calculated values. The value decreased significantly.

[0050] This invention makes a judgment by combining two dimensions: slope difference. The significant change indicates that the sensor condition has changed from dry to wet; the goodness of fit R 2 The significant decrease further confirms that this is an ongoing dynamic process (i.e., liquid is seeping in), rather than a static result. This multi-feature combined judgment method greatly improves the accuracy of leak detection and the ability to capture dynamic processes.

[0051] Furthermore, the method for extracting robust features in step S3 also includes:

[0052] The difference in Received Signal Strength Indication (RSSI) between dual RFID tag antennas is introduced as an auxiliary judgment feature in addition to the phase fitting feature. The calculation expression for the difference in Received Signal Strength Indication is as follows:

[0053] ;

[0054] in, It is the signal strength value of tag A received by the reader;

[0055] It is the signal strength value of tag B received by the reader;

[0056] In a dry state, due to the close arrangement of the dual RFID tag antennas, there is asymmetrical electric field coupling between the two RFID tag antennas. The electromagnetic field generated by one RFID tag antenna will affect the other RFID tag antenna. The asymmetry in the physical structure causes the signal of one RFID tag to be partially suppressed, resulting in different signal power transmitted back to the reader by the two RFID tags. This manifests as a stable and non-zero RSSI difference at the reader end, and the RSSI difference is used as the reference state for the normal operation of the dual tag sensor.

[0057] When a liquid leak occurs, the liquid is absorbed by the absorbing medium, which changes the dielectric constant of the coupling medium between the two RFID tags and the antenna impedance of the RFID tags themselves. This weakens the electric field coupling strength between the two RFID tags, thereby weakening the original signal suppression effect. As a result, the RSSI difference between the two RFID tags decreases rapidly. When the RSSI difference is observed to narrow rapidly from a stable value, it is determined that the liquid is soaking the dual-tag sensor.

[0058] Preferably, this feature is compared with the aforementioned phase fitting feature (slope difference). and goodness of fit By combining these methods, a more reliable multi-dimensional decision-making system can be built, effectively distinguishing between leakage events and environmental interference, further enhancing the robustness of detection, and thus achieving high-precision leak detection.

[0059] Furthermore, the method for comprehensively analyzing feature vectors using a decision tree classification model in step S4 includes:

[0060] The input feature vector is constructed by including the phase-frequency slope difference Δk and the goodness-of-fit R. 2 And a decision tree classification model for RSSI difference, automatically learning and constructing judgment rules to optimally combine the thresholds of various feature vectors; judging and classifying different states of normal, interference, and leakage, including:

[0061] Determine whether the phase-frequency slope difference Δk has significantly deviated from its stable reference value, i.e. whether liquid leakage has occurred. Preferably, the threshold for Δk is 0.0015 rad / Hz.

[0062] Determine the goodness of fit R 2 Whether it decreases significantly, i.e. whether the liquid leakage process occurs dynamically, preferably, the goodness of fit R 2 The threshold is 0.8;

[0063] To determine whether the RSSI difference decreases rapidly due to the weakening of the electric field coupling strength between the two RFID tag antennas, i.e. whether liquid is wetting the dual-tag sensor, preferably, the threshold for the RSSI difference is 3dB.

[0064] A single feature can only reflect one aspect of a leak, while comprehensive analysis and judgment can significantly improve the robustness of detection.

[0065] This decision tree classification model does not require a large amount of training data. Experiments have shown that by training with a small number of clearly labeled leak and non-leak event data, the decision tree classification model can quickly converge and grasp the key classification rules, ultimately achieving a TPR of up to 92% and an FPR of 2%.

[0066] The present invention also provides a dual RFID leakage detection device based on multi-frequency feature fusion, for performing the dual RFID leakage detection method based on multi-frequency feature fusion as described above, comprising:

[0067] Leakage detection sensor module: A pair of UHF RFID tags are fixed in parallel and close to each other as a dual-tag sensor, and a liquid absorption medium is set between the two RFID tags to capture leaked liquid and amplify its effects.

[0068] Leakage sensing model module: used to construct a leakage sensing model between two RFID tags based on the mutual inductance coupling effect between the antenna circuits of adjacent RFID tags. The leakage sensing model is characterized as a mathematical model that causes a change in the signal phase of the RFID tag itself and the adjacent RFID tags when the impedance of one RFID tag changes.

[0069] Multi-frequency phase fitting feature extraction module: This module utilizes the frequency hopping characteristic of the EPC Gen2 protocol to perform unified analysis of the phases of multiple RFID tag signals at multiple frequencies, extracting robust features, including: features based on signal phase. With operating frequency A linear relationship exists: , where the slope represents the rate of phase change with frequency; b is the intercept of the phase-frequency curve; multiple data points acquired within one frequency hopping cycle... Linear fitting is performed on the data points to obtain the key feature parameter, slope. and goodness of fit According to the slope and / or goodness of fit Determine whether the dielectric properties of the dual-tag sensor have been altered due to liquid leakage;

[0070] Decision model and comprehensive judgment module: used to fuse the extracted multiple features, and use a decision tree classification model to comprehensively analyze the feature vectors, effectively distinguishing between real leakage, environmental interference and sensor static changes.

[0071] Furthermore, the multi-frequency phase fitting feature extraction module includes:

[0072] Slope k judgment unit: Used to determine the rate k of phase change of the phase-frequency curve with frequency as an indicator of leakage status, slope. ,in, It is a component determined by the signal propagation distance. The component is determined by the RFID tag hardware; two RFID tags are fixedly placed. The components are stable and equal. Component common mode; slope values ​​for two RFID tags. and Perform a difference operation. and The difference between them is slope difference The difference in impedance components between two RFID tags The decision was made to place the product in a dry state. It provides a stable reference value, thus eliminating the influence of propagation distance and reader hardware. When liquid wets the dual-tag sensor, it asymmetrically changes the dielectric constant around the two tag antennas, causing different changes in the equivalent input impedance of the two tag antennas, leading to... and The changes are no longer synchronized, causing the slope difference to... Deviation from its baseline value in the dry state; by monitoring the slope difference between the two labels. Whether there is a significant deviation from a stable reference value determines whether the dielectric properties of the dual-tag sensor have changed due to liquid leakage;

[0073] Goodness of fit Judgment unit: used for goodness-of-fit assessment The goodness-of-fit is a characteristic that reflects the degree of fit between data points and the fitted linear model. As an indicator of whether the leakage process is occurring dynamically; the goodness of fit The calculation methods include: obtaining a straight line by fitting using the least squares method. ;in It is the model prediction value of the signal phase. It is the actual frequency; calculation of observation data Relative to its average value Sum of squares of total variation: ; Calculate the actual observed values and model predictions Sum of squares of the differences between them (SSE): Based on SST and SSE, the goodness of fit is calculated. : ;like If the value drops significantly, it is determined that a leakage process is occurring dynamically.

[0074] The present invention also provides a computer-readable storage medium having a computer program stored thereon, wherein the program, when executed by a processor, implements the steps of the dual RFID leakage detection method based on multi-frequency feature fusion as described above.

[0075] The present invention also provides a computer device, the computer device including a memory, a processor and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, it implements the steps of the dual RFID leakage detection method based on multi-frequency feature fusion as described above.

[0076] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0077] The dual RFID leak detection method based on multi-frequency feature fusion provided by this invention does not rely on the absolute value of a single tag signal. Based on the coupling effect of dual RFID tags, it utilizes the coupling characteristics of adjacent RFID tags to sense changes in the medium caused by liquid leakage, monitors the relative relationship between the signals of two closely adjacent tags, effectively distinguishes leak events from environmental interference, and significantly reduces the false alarm rate. Furthermore, it combines RFID frequency hopping mechanism to obtain multi-frequency phase data, extracts features through linear fitting, and achieves high-precision detection of the leak status. Moreover, it requires no additional hardware, has high sensitivity, strong real-time performance, good applicability, and has broad prospects for promotion and application. Attached Figure Description

[0078] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the invention.

[0079] In the attached diagram:

[0080] Figure 1 This is a design diagram of dual RFID tags according to an embodiment of the present invention;

[0081] Figure 2 This is an equivalent circuit model diagram of the leakage sensing model according to an embodiment of the present invention;

[0082] Figure 3 This is an analysis diagram of the electric field model of a pair of parallel and adjacent RFID tags according to an embodiment of the present invention.

[0083] Figure 4 This is a diagram showing the model judgment results of the decision tree classification model in an embodiment of the present invention;

[0084] Figure 5 This is a graph showing the phase and frequency changes according to an embodiment of the present invention.

[0085] Figure 6 This is an architecture diagram of the decision tree classification model according to an embodiment of the present invention;

[0086] Figure 7 This is a flowchart of a dual RFID leakage detection method based on multi-frequency feature fusion according to an embodiment of the present invention;

[0087] Figure 8 This is a schematic diagram of the configuration of a computer device according to an embodiment of the present invention. Detailed Implementation

[0088] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numerals in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this disclosure. Rather, they are merely examples of apparatuses and products consistent with some aspects of this disclosure as detailed in the appended claims.

[0089] The terminology used in this disclosure is for the purpose of describing particular embodiments only and is not intended to be limiting of the disclosure. The singular forms “a,” “the,” and “the” as used in this disclosure and the appended claims are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used herein refers to and includes any and all possible combinations of one or more of the associated listed items.

[0090] It should be understood that although the terms first, second, third, etc., may be used in this disclosure to describe various information, such information should not be limited to these terms. These terms are used only to distinguish information of the same type from one another. For example, without departing from the scope of this disclosure, first information may also be referred to as second information, and similarly, second information may also be referred to as first information. Depending on the context, the word "if" as used herein may be interpreted as "when," "when," or "in response to determination."

[0091] The embodiments of the present invention will be described in further detail below.

[0092] This invention provides a dual RFID leakage detection method based on multi-frequency feature fusion, see [link to relevant documentation]. Figure 7 As shown, it includes the following steps:

[0093] S1. A pair of ultra-high frequency (UHF) RFID tags are fixed in parallel and close together as a dual-tag sensor (e.g. Figure 1 As shown in the figure, a liquid-absorbing medium is placed between the two RFID tags to capture the leaked liquid and amplify its impact;

[0094] S2. Based on the mutual inductance coupling effect between the antenna circuits of adjacent RFID tags, a leakage sensing model is constructed between the two RFID tags. The leakage sensing model is characterized by a mathematical model that causes a change in the signal phase of the RFID tag itself and its neighboring RFID tags when the impedance of one RFID tag changes. Step S2 includes:

[0095] The interaction between two adjacent RFID tags, A and B, can be simplified into an equivalent circuit model (e.g., Figure 2 As shown), the current I flowing through tag A and tag B is... A and I B The relational equation followed is: ,

[0096] (1)

[0097] In equation (1), , These are the induced source voltages on label A and label B, respectively;

[0098] , The chip impedances are for labels A and B, respectively.

[0099] , These are the impedances of the antennas of tag A and tag B, respectively.

[0100] , These are the mutual impedances between label A and label B, respectively.

[0101] Solving equation (1), we obtain the current I of tag A and tag B. A and I B ;

[0102] The current of any tag is determined by all impedance parameters in the system. Changing any impedance value (such as...) This will simultaneously affect and , and The relationship with impedance is:

[0103] ,

[0104] (2)

[0105] In equation (2), and These represent the total impedances of label A and label B, respectively. and The calculation expressions are as follows:

[0106] ,

[0107] (3)

[0108] The signal phase of an RFID tag is determined by the complex amplitude of the antenna current. The current phase values ​​for tag A and tag B are as follows:

[0109] ,

[0110] (4)

[0111] When an external disturbance is applied to one of the tags, causing changes in the tag's own impedance and mutual impedance, the current phase of both tags will undergo predictable changes.

[0112] In this embodiment, when an external disturbance is applied to tag B, causing its own impedance to be affected... and mutual impedance When this changes, it will cause the current phase of the two tags to change. and All of these undergo predictable changes, allowing physical events to be inferred by measuring the phase matrix;

[0113] S3. Utilizing the frequency hopping feature in the EPC Gen2 protocol, perform unified analysis on the signal phases of multiple RFID tags at multiple frequencies to extract robust features, including: based on signal phase... With operating frequency A linear relationship exists: , where the slope represents the rate of phase change with frequency; b is the intercept of the phase-frequency curve, as shown in the figure. Figure 5 As shown; multiple data collected within one frequency hopping cycle Linear fitting is performed on the data points to obtain the key feature parameter, slope. and goodness of fit According to the slope and / or goodness of fit Determine whether the dielectric properties of the dual-tag sensor have been altered due to liquid leakage;

[0114] Methods for determining whether the dielectric properties of a dual-tag sensor have changed due to liquid leakage based on the slope k include:

[0115] The rate of change of phase with frequency (linear rate of change) k of the phase-frequency curve is used as an indicator of leakage status, and the slope is... ,in, It is a component determined by the signal propagation distance. It is a component determined by the RFID tag hardware (especially the antenna equivalent input impedance);

[0116] In the dual-tag sensor of this embodiment, two tags are fixedly placed, therefore their The components are stable and nearly equal. Errors introduced by the reader hardware. It is also common mode. Therefore,

[0117] Two RFID tags were fixed in place. The components are stable and equal. Component common mode; slope values ​​for two RFID tags. and Perform a difference operation. and The difference between them is slope difference The difference in impedance components between two RFID tags The decision was made to place the product in a dry state. It is a stable benchmark value, thus eliminating the influence of propagation distance and reader hardware;

[0118] When a liquid wets a dual-tag sensor, it asymmetrically changes the dielectric constant around the two tag antennas, causing different changes in the equivalent input impedance of the two tag antennas, resulting in... and The changes are no longer synchronized, causing the slope difference to... Deviation from its baseline value in the dry state; by monitoring the slope difference between the two labels. Whether there is a significant deviation from a stable reference value determines whether the dielectric properties of the dual-tag sensor have changed due to liquid leakage.

[0119] Based on goodness of fit Methods for determining whether liquid leakage has altered the dielectric properties of a dual-tag sensor include:

[0120] Based on goodness of fit The goodness-of-fit is a characteristic that reflects the degree of fit between data points and the fitted linear model. As a marker to determine whether a leakage process is occurring dynamically;

[0121] The goodness of fit The calculation methods include: obtaining a straight line by fitting using the least squares method. ;in It is the model prediction value of the signal phase. It is the actual frequency;

[0122] Calculate observation data Relative to its average value Sum of squares of total variation:

[0123] ;

[0124] Calculate the actual observed values and model predictions The sum of squares (SSE) of the differences (i.e., residuals) between them;

[0125] ;

[0126] The goodness of fit is calculated based on SST and SSE. :

[0127] ;

[0128] like If the value drops significantly, it is determined that a leakage process is occurring dynamically.

[0129] Under stable conditions of dryness or liquid saturation, the impedance of an RFID tag is constant, and the data collected within a complete frequency hopping cycle... The data points can well conform to a single linear relationship, and the calculated values ​​are... The value is close to 1.

[0130] Liquid leakage is a dynamic process, with the amount of liquid continuously increasing. During the brief timeframe of frequency hopping measurements, the tag's impedance changes continuously. Therefore, the phase measured at different frequencies actually corresponds to the state at different times and under different impedances. When linearly fitting these data points corresponding to dynamically changing states, they will no longer lie strictly on a straight line, causing the data points to exhibit a discrete or nonlinear trend. This deteriorates the linear fitting effect, leading to errors in the calculated values. The value decreased significantly.

[0131] This embodiment makes a judgment by combining two dimensions: slope difference. The significant change indicates that the sensor condition has changed from dry to wet; the goodness of fit R 2The significant decrease further confirms that this is an ongoing dynamic process (i.e., liquid is seeping in), rather than a static result. This multi-feature combined judgment method greatly improves the accuracy of leak detection and the ability to capture dynamic processes.

[0132] The method for extracting robust features in this embodiment also includes:

[0133] The difference in Received Signal Strength Indication (RSSI) between dual RFID tag antennas is introduced as an auxiliary judgment feature in addition to the phase fitting feature. The calculation expression for the difference in Received Signal Strength Indication is as follows:

[0134] ;

[0135] in, It is the signal strength value of tag A received by the reader;

[0136] It is the signal strength value of tag B received by the reader;

[0137] In a dry state, due to the close arrangement of the dual RFID tag antennas, there is asymmetric electric field coupling between the two RFID tag antennas (e.g., Figure 3 As shown, the electromagnetic field generated by one RFID tag antenna will affect the other RFID tag antenna. The asymmetry in the physical structure causes the signal of one RFID tag to be partially suppressed, resulting in different signal power transmitted back to the reader by the two RFID tags. This results in a stable and non-zero RSSI difference at the reader end, which is used as the reference state for the normal operation of the dual tag sensor.

[0138] When a liquid leak occurs, the liquid is absorbed by the absorbing medium, which changes the dielectric constant of the coupling medium between the two RFID tags and the antenna impedance of the RFID tags themselves. This weakens the electric field coupling strength between the two RFID tags, thereby weakening the original signal suppression effect. As a result, the RSSI difference between the two RFID tags decreases rapidly. When the RSSI difference is observed to narrow rapidly from a stable value, it is determined that the liquid is soaking the dual-tag sensor.

[0139] The RSSI difference feature is compared with the aforementioned phase fitting feature (slope difference). and goodness of fit By combining these methods, a more reliable multi-dimensional decision-making system can be built, effectively distinguishing between leakage events and environmental interference, further enhancing the robustness of detection, and thus achieving high-precision leak detection.

[0140] S4. The extracted features are fused together, and a decision tree classification model (such as...) is used. Figure 6As shown, a comprehensive analysis of the feature vectors is performed to effectively distinguish between real leakage, environmental interference, and static changes in the sensor.

[0141] A single feature can only reflect one aspect of a leak, while comprehensive analysis and judgment significantly improve the robustness of detection.

[0142] Methods for comprehensively analyzing feature vectors using decision tree classification models include:

[0143] The input feature vector is constructed by including the phase-frequency slope difference Δk and the goodness-of-fit R. 2 And a decision tree classification model for RSSI difference, automatically learning and constructing judgment rules to optimally combine the thresholds of various feature vectors; judging and classifying different states of normal, interference, and leakage, including:

[0144] To determine whether the phase-frequency slope difference Δk has significantly deviated from its stable reference value, i.e. whether liquid leakage has occurred, the threshold for Δk is 0.0015 rad / Hz.

[0145] Determine the goodness of fit R 2 Whether it decreases significantly, i.e. whether the liquid leakage process occurs dynamically, and the goodness of fit R. 2 The threshold is 0.8;

[0146] To determine whether the RSSI difference decreases rapidly due to the weakening of the electric field coupling strength between the two RFID tag antennas, i.e. whether liquid is wetting the dual-tag sensor, the threshold for the RSSI difference is 3dB.

[0147] The decision tree classification model in this embodiment does not require a large amount of training data. Through small-sample training (i.e., training with a small amount of clearly labeled leak and non-leak event data), the model can quickly converge and grasp the key classification rules, ultimately achieving a TPR of up to 92% and an FPR of 2% (e.g., ...). Figure 4 (As shown).

[0148] This invention also provides a dual RFID leakage detection device based on multi-frequency feature fusion, used to perform the dual RFID leakage detection method based on multi-frequency feature fusion as described above, including:

[0149] Leakage detection sensor module: A pair of UHF RFID tags are fixed in parallel and close to each other as a dual-tag sensor, and a liquid absorption medium is set between the two RFID tags to capture leaked liquid and amplify its effects.

[0150] Leakage sensing model module: used to construct a leakage sensing model between two RFID tags based on the mutual inductance coupling effect between the antenna circuits of adjacent RFID tags. The leakage sensing model is characterized as a mathematical model that causes a change in the signal phase of the RFID tag itself and the adjacent RFID tags when the impedance of one RFID tag changes.

[0151] Multi-frequency phase fitting feature extraction module: This module utilizes the frequency hopping characteristic of the EPC Gen2 protocol to perform unified analysis of the phases of multiple RFID tag signals at multiple frequencies, extracting robust features, including: features based on signal phase. With operating frequency A linear relationship exists: , where the slope represents the rate of phase change with frequency; b is the intercept of the phase-frequency curve; multiple data points acquired within one frequency hopping cycle... Linear fitting is performed on the data points to obtain the key feature parameter, slope. and goodness of fit According to the slope and / or goodness of fit Determine whether the dielectric properties of the dual-tag sensor have been altered due to liquid leakage; the multi-frequency phase fitting feature extraction module includes:

[0152] Slope k judgment unit: Used to determine the rate k of phase change of the phase-frequency curve with frequency as an indicator of leakage status, slope. ,in, It is a component determined by the signal propagation distance. The component is determined by the RFID tag hardware; two RFID tags are fixedly placed. The components are stable and equal. Component common mode; slope values ​​for two RFID tags. and Perform a difference operation. and The difference between them is slope difference The difference in impedance components between two RFID tags The decision was made to place the product in a dry state. It provides a stable reference value, thus eliminating the influence of propagation distance and reader hardware. When liquid wets the dual-tag sensor, it asymmetrically changes the dielectric constant around the two tag antennas, causing different changes in the equivalent input impedance of the two tag antennas, leading to... and The changes are no longer synchronized, causing the slope difference to... Deviation from its baseline value in the dry state; by monitoring the slope difference between the two labels. Whether there is a significant deviation from a stable reference value determines whether the dielectric properties of the dual-tag sensor have changed due to liquid leakage;

[0153] Goodness of fit Judgment unit: used for goodness-of-fit assessment The goodness-of-fit is a characteristic that reflects the degree of fit between data points and the fitted linear model. As an indicator of whether the leakage process is occurring dynamically; the goodness of fit The calculation methods include: obtaining a straight line by fitting using the least squares method. ;in It is the model prediction value of the signal phase. It is the actual frequency; calculation of observation data Relative to its average value Sum of squares of total variation: ; Calculate the actual observed values and model predictions Sum of squares of the differences between them (SSE): Based on SST and SSE, the goodness of fit is calculated. : ;like If the value drops significantly, it is determined that a leakage process is occurring dynamically.

[0154] Decision model and comprehensive judgment module: used to fuse the extracted multiple features, and use a decision tree classification model to comprehensively analyze the feature vectors, effectively distinguishing between real leakage, environmental interference and sensor static changes.

[0155] This invention also provides a computer device. Figure 8 This is a schematic diagram of the structure of a computer device provided in an embodiment of the present invention; see the accompanying drawings. Figure 8 As shown, the computer device includes: an input system 23, an output system 24, a memory 22, and a processor 21; the memory 22 is used to store one or more programs; when the one or more programs are executed by the one or more processors 21, the one or more processors 21 implement the dual RFID leakage detection method based on multi-frequency feature fusion as provided in the above embodiment; wherein the input system 23, the output system 24, the memory 22, and the processor 21 can be connected via a bus or other means. Figure 8 Taking the example of a connection between China and Israel via a bus.

[0156] The memory 22, as a read / write storage medium for a computing device, can be used to store software programs and computer-executable programs, such as the program instructions corresponding to the dual RFID leakage detection method based on multi-frequency feature fusion described in this embodiment of the invention. The memory 22 may mainly include a program storage area and a data storage area. The program storage area may store the operating system and at least one application program required for a function; the data storage area may store data created based on the use of the device. Furthermore, the memory 22 may include high-speed random access memory and non-volatile memory, such as at least one disk storage device, flash memory device, or other non-volatile solid-state storage device. In some instances, the memory 22 may further include memory remotely located relative to the processor 21, and these remote memories can be connected to the device via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0157] The input system 23 can be used to receive input digital or character information, and generate key signal inputs related to user settings and function control of the device; the output system 24 may include display devices such as a display screen.

[0158] The processor 21 executes various functional applications and data processing of the device by running software programs, instructions and modules stored in the memory 22, thereby realizing the above-mentioned dual RFID leakage detection method based on multi-frequency feature fusion.

[0159] The computer equipment provided above can be used to execute the dual RFID leakage detection method based on multi-frequency feature fusion provided in the above embodiments, and has corresponding functions and beneficial effects.

[0160] This invention also provides a storage medium containing computer-executable instructions, which, when executed by a computer processor, are used to perform the dual RFID leakage detection method based on multi-frequency feature fusion as provided in the above embodiments. The storage medium can be any type of memory device or storage device, including: mounting media such as CD-ROM, floppy disk, or magnetic tape systems; computer system memory or random access memory such as DRAM, DDRRAM, SRAM, EDO RAM, Rambus RAM, etc.; non-volatile memory such as flash memory, magnetic media (e.g., hard disk or optical storage); registers or other similar types of memory elements; the storage medium may also include other types of memory or combinations thereof; furthermore, the storage medium may reside in a first computer system in which the program is executed, or it may reside in a different second computer system connected to the first computer system via a network (such as the Internet); the second computer system can provide program instructions to the first computer for execution. The storage medium includes two or more storage media that can reside in different locations (e.g., in different computer systems connected via a network). The storage medium can store program instructions (e.g., specifically implemented as a computer program) executable by one or more processors.

[0161] Of course, the computer-executable instructions provided in the embodiments of the present invention are not limited to the dual RFID leakage detection method based on multi-frequency feature fusion as described in the above embodiments, but can also perform related operations in the dual RFID leakage detection method based on multi-frequency feature fusion provided in any embodiment of the present invention.

[0162] The technical solution of the present invention has been described in conjunction with preferred embodiments. However, it will be readily understood by those skilled in the art that the scope of protection of the present invention is obviously not limited to these specific embodiments. Without departing from the principles of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions resulting from these changes or substitutions will all fall within the scope of protection of the present invention.

[0163] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A dual RFID leakage detection method based on multi-frequency feature fusion, characterized in that, Includes the following steps: S1. A pair of UHF RFID tags are fixed in parallel and close to each other as a dual-tag sensor. A liquid absorption medium is set between the two RFID tags to capture the leaked liquid and amplify its effect. S2. Based on the mutual inductance coupling effect between the antenna circuits of adjacent RFID tags, a leakage sensing model between the two RFID tags is constructed. The leakage sensing model is characterized as a mathematical model that causes a change in the signal phase of the RFID tag itself and the adjacent RFID tags when the impedance of one RFID tag changes. S3. Utilizing the frequency hopping characteristic in the EPC Gen2 protocol, perform unified analysis on the signal phases of multiple RFID tags at multiple frequencies to extract robust features, including: based on signal phase. With operating frequency A linear relationship exists: The slope It represents the rate at which the phase changes with frequency; b is the intercept of the phase-frequency curve; it is obtained by collecting multiple data points within one frequency hopping cycle. Linear fitting is performed on the data points to obtain the key feature parameter, slope. and goodness of fit According to the slope and / or goodness of fit Determine whether the dielectric properties of the dual-tag sensor have been altered due to liquid leakage; S4. The extracted features are fused together, and a decision tree classification model is used to comprehensively analyze the feature vectors to effectively distinguish between real leakage, environmental interference, and static changes of the sensor.

2. The dual RFID leakage detection method based on multi-frequency feature fusion according to claim 1, characterized in that, Step S2 includes: The interaction between two adjacent RFID tags, A and B, can be simplified into an equivalent circuit model. The current I flowing through tag A and tag B is... A and I B The relational equation followed is: , （1） In equation (1), , These are the induced source voltages on label A and label B, respectively; , The chip impedances are for labels A and B, respectively. , These are the impedances of the antennas of tag A and tag B, respectively. , These are the mutual impedances between label A and label B, respectively. Solving equation (1), we obtain the current I of tag A and tag B. A and I B ; The current of any tag is determined by all impedance parameters in the system; changing any impedance value will simultaneously affect... and , and The relationship with impedance is: ， （2） In equation (2), and These represent the total impedances of label A and label B, respectively. and The calculation expressions are as follows: ， （3） The signal phase of an RFID tag is determined by the complex amplitude of the antenna current. The current phase values ​​for tag A and tag B are as follows: ， （4） When an external disturbance is applied to one of the tags, causing changes in the tag's own impedance and mutual impedance, the current phase of both tags will undergo predictable changes.

3. The dual RFID leakage detection method based on multi-frequency feature fusion according to claim 1, characterized in that, The method for determining whether the dielectric properties of the dual-tag sensor have changed due to liquid leakage based on the slope k in step S3 includes: The rate of change k of the phase versus frequency curve is used as an indicator of leakage status, and the slope is... ,in, It is a component determined by the signal propagation distance. The weight is determined by the RFID tag hardware; Two RFID tags were fixed in place. The components are stable and equal. Component common mode; slope values ​​for two RFID tags. and Perform a difference operation. and The difference between them is slope difference The difference in impedance components between two RFID tags The decision was made to place the product in a dry state. It is a stable benchmark value, thus eliminating the influence of propagation distance and reader hardware; When a liquid wets a dual-tag sensor, it asymmetrically changes the dielectric constant around the two tag antennas, causing different changes in the equivalent input impedance of the two tag antennas, resulting in... and The changes are no longer synchronized, causing the slope difference to... Deviation from its baseline value in the dry state; by monitoring the slope difference between the two labels. Whether there is a significant deviation from a stable reference value determines whether the dielectric properties of the dual-tag sensor have changed due to liquid leakage.

4. The dual RFID leakage detection method based on multi-frequency feature fusion according to claim 1, characterized in that, The S3 step is based on the goodness of fit. Methods for determining whether liquid leakage has altered the dielectric properties of a dual-tag sensor include: Based on goodness of fit The goodness-of-fit is a characteristic that reflects the degree of fit between data points and the fitted linear model. As a marker to determine whether a leakage process is occurring dynamically; The goodness of fit The calculation methods include: obtaining a straight line by fitting using the least squares method. ;in It is the model prediction value of the signal phase. It is the actual frequency; Calculate observation data Relative to its average value Sum of squares of total variation: ； Calculate the actual observed values and model predictions The sum of squares of the differences between them (SSE); ； The goodness of fit is calculated based on SST and SSE. : ； like If the value drops significantly, it is determined that a leakage process is occurring dynamically.

5. The dual RFID leakage detection method based on multi-frequency feature fusion according to claim 4, characterized in that, The method for extracting robust features in step S3 also includes: The difference in received signal strength indication from dual RFID tag antennas is introduced as an auxiliary judgment feature in addition to the phase fitting feature. The calculation expression for the difference in received signal strength indication is as follows: ； in, It is the signal strength value of tag A received by the reader; It is the signal strength value of tag B received by the reader; In a dry state, due to the close arrangement of the dual RFID tag antennas, there is asymmetrical electric field coupling between the two RFID tag antennas. The electromagnetic field generated by one RFID tag antenna will affect the other RFID tag antenna. The asymmetry in the physical structure causes the signal of one RFID tag to be partially suppressed, resulting in different signal power transmitted back to the reader by the two RFID tags. This manifests as a stable and non-zero RSSI difference at the reader end, and the RSSI difference is used as the reference state for the normal operation of the dual tag sensor. When a liquid leak occurs, the liquid is absorbed by the absorbing medium, which changes the dielectric constant of the coupling medium between the two RFID tags and the antenna impedance of the RFID tags themselves. This weakens the electric field coupling strength between the two RFID tags, thereby weakening the original signal suppression effect. As a result, the RSSI difference between the two RFID tags decreases rapidly. When the RSSI difference is observed to narrow rapidly from a stable value, it is determined that the liquid is soaking the dual-tag sensor.

6. The dual RFID leakage detection method based on multi-frequency feature fusion according to claim 5, characterized in that, The method for comprehensively analyzing feature vectors using a decision tree classification model in step S4 includes: The input feature vector is constructed by including the phase-frequency slope difference Δk and the goodness-of-fit R. 2 And a decision tree classification model for RSSI difference, automatically learning and constructing judgment rules to optimally combine the thresholds of various feature vectors; judging and classifying different states of normal, interference, and leakage, including: Determine whether the phase-frequency slope difference Δk has shifted significantly from its stable reference value, i.e., whether a liquid leak has been detected; Determine the goodness of fit R 2 Whether there is a significant decrease, i.e. whether the liquid leakage process occurs dynamically; Determine whether the RSSI difference decreases rapidly due to the weakening of the electric field coupling strength between the two RFID tag antennas, i.e., whether liquid is wetting the dual tag sensor.

7. A dual RFID leakage detection device based on multi-frequency feature fusion, used to execute the dual RFID leakage detection method based on multi-frequency feature fusion as described in any one of claims 1-6, characterized in that, include: Leakage detection sensor module: A pair of UHF RFID tags are fixed in parallel and close to each other as a dual-tag sensor, and a liquid absorption medium is set between the two RFID tags to capture leaked liquid and amplify its effects. Leakage sensing model module: used to construct a leakage sensing model between two RFID tags based on the mutual inductance coupling effect between the antenna circuits of adjacent RFID tags. The leakage sensing model is characterized as a mathematical model that causes a change in the signal phase of the RFID tag itself and the adjacent RFID tags when the impedance of one RFID tag changes. Multi-frequency phase fitting feature extraction module: This module utilizes the frequency hopping characteristic of the EPC Gen2 protocol to perform unified analysis of the phases of multiple RFID tag signals at multiple frequencies, extracting robust features, including: features based on signal phase. With operating frequency A linear relationship exists: The slope It represents the rate at which the phase changes with frequency; b is the intercept of the phase-frequency curve; it is obtained by collecting multiple data points within one frequency hopping cycle. Linear fitting is performed on the data points to obtain the key feature parameter, slope. and goodness of fit According to the slope and / or goodness of fit Determine whether the dielectric properties of the dual-tag sensor have been altered due to liquid leakage; Decision model and comprehensive judgment module: used to fuse the extracted multiple features, and use a decision tree classification model to comprehensively analyze the feature vectors, effectively distinguishing between real leakage, environmental interference and sensor static changes.

8. The dual RFID leakage detection device based on multi-frequency feature fusion according to claim 7, characterized in that, The multi-frequency phase fitting feature extraction module includes: Slope k judgment unit: Used to determine the rate k of phase change of the phase-frequency curve with frequency as an indicator of leakage status, slope. ,in, It is a component determined by the signal propagation distance. The component is determined by the RFID tag hardware; two RFID tags are fixedly placed. The components are stable and equal. Component common mode; slope values ​​for two RFID tags. and Perform a difference operation. and The difference between them is slope difference The difference in impedance components between two RFID tags The decision was made to place the product in a dry state. It provides a stable reference value, thus eliminating the influence of propagation distance and reader hardware. When liquid wets the dual-tag sensor, it asymmetrically changes the dielectric constant around the two tag antennas, causing different changes in the equivalent input impedance of the two tag antennas, leading to... and The changes are no longer synchronized, causing the slope difference to... Deviation from its baseline value in the dry state; by monitoring the slope difference between the two labels. Whether there is a significant deviation from a stable reference value determines whether the dielectric properties of the dual-tag sensor have changed due to liquid leakage; Goodness of fit Judgment unit: used for goodness-of-fit assessment The goodness-of-fit is a characteristic that reflects the degree of fit between data points and the fitted linear model. As an indicator of whether the leakage process is occurring dynamically; the goodness of fit The calculation methods include: obtaining a straight line by fitting using the least squares method. ;in It is the model prediction value of the signal phase. It is the actual frequency; calculation of observation data Relative to its average value Sum of squares of total variation: ; Calculate the actual observed values and model predictions Sum of squares of the differences between them (SSE): Based on SST and SSE, the goodness of fit is calculated. : ;like If the value drops significantly, it is determined that a leakage process is occurring dynamically.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the steps of the dual RFID leakage detection method based on multi-frequency feature fusion as described in any one of claims 1-6.

10. A computer device, the computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the steps of the dual RFID leakage detection method based on multi-frequency feature fusion as described in any one of claims 1-6.

Images