A motion-robust temperature monitoring system and method based on RFID passive multi-frequency sensing

Abstract

This invention discloses a motion-robust temperature monitoring system and method based on RFID passive multi-frequency sensing, comprising: a passive temperature sensor for receiving radio frequency signals emitted by a temperature sensing integrated unit; a temperature sensing integrated unit for generating multi-frequency radio frequency signals to activate the passive temperature sensor and acquiring the backscattered radio frequency signals from the passive temperature sensor in real time; and a data processing platform for acquiring the radio frequency signals in real time, performing subcarrier synchronization, random phase offset elimination, and multi-frequency differential processing on the radio frequency signals, extracting motion-independent temperature fingerprints, and performing temperature prediction to obtain the temperature of the target object. This invention combines motion-independent temperature features corresponding to multiple frequency pairs into a temperature fingerprint, and uses a temperature prediction model to enhance and regress the temperature fingerprint, improving the distinguishability of temperature fingerprints under similar temperatures, thereby improving the accuracy of temperature prediction.

Description

Technical Field

[0001] This invention belongs to the field of RFID passive sensing technology, specifically relating to a motion-robust temperature monitoring system and method based on RFID passive multi-frequency sensing. Background Technology

[0002] Temperature is a crucial fundamental indicator in environmental monitoring, industrial production, cold chain logistics, warehouse management, and human health monitoring. Continuous, stable, and low-cost temperature monitoring of target objects can provide important data for tracking the status of goods, managing equipment safety, and issuing early warnings of health anomalies. However, in real-world applications, the target object, sensing device, or surrounding environment are often in a dynamic state of change. For example, the movement of objects on a conveyor belt, the swaying of goods during cold chain transportation, changes in patient positioning, and movement of people in the environment can all cause changes in the wireless signal propagation path and multipath environment, thus affecting the stability of temperature sensing.

[0003] Existing temperature monitoring methods mainly include contact temperature sensors, infrared temperature measurement devices, and wireless signal-based temperature sensing methods. Contact temperature sensors typically require additional dedicated sensors and rely on batteries or external power supplies, resulting in high costs, complex maintenance, and limited battery life. While infrared temperature measurement devices can achieve non-contact temperature measurement, they usually require good line-of-sight conditions and high alignment accuracy, and are prone to errors when the target moves or is obstructed. Existing RFID-based temperature sensing methods have advantages such as low cost, passivity, and ease of deployment, but temperature-induced signal changes are easily affected by device movement and environmental motion, leading to a decrease in temperature measurement accuracy.

[0004] In addition, some methods eliminate motion interference by deploying reference tags, but this increases the number of tags and deployment complexity, making it difficult to directly utilize individual commercial RFID tags that are already widely deployed in existing IoT scenarios.

[0005] Therefore, based on the above considerations, there is an urgent need for a low-cost, passive RFID temperature monitoring system and method that does not require reference tags and can work stably in motion interference scenarios, so as to achieve robust temperature sensing in complex dynamic environments. Summary of the Invention

[0006] To address the shortcomings of the existing technologies, the present invention aims to provide a motion-robust temperature monitoring system and method based on RFID passive multi-frequency sensing, in order to solve the problems of existing technologies such as reliance on active dedicated sensors, high deployment costs, difficulty in utilizing existing single commercial RFID tags, and decreased temperature measurement accuracy under interference from equipment movement and environmental movement.

[0007] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0008] The present invention provides a motion-robust temperature monitoring system based on RFID passive multi-frequency sensing, comprising: a passive temperature sensor, an integrated temperature sensing unit, and a data processing platform;

[0009] A passive temperature sensor is used to receive the wireless radio frequency signal emitted by the temperature sensing all-in-one machine, and then transmit the received wireless radio frequency signal to the temperature sensing all-in-one machine after backscattering modulation.

[0010] The temperature sensing all-in-one machine is used to generate multi-frequency wireless radio frequency signals to activate the passive temperature sensor, and to collect the backscattered wireless radio frequency signals of the passive temperature sensor in real time and send them to the data processing platform.

[0011] The data processing platform is used to acquire the wireless radio frequency signal sent by the temperature sensing all-in-one machine in real time, perform subcarrier synchronization, random phase offset elimination and multi-frequency differential processing on the wireless radio frequency signal, extract motion-independent temperature fingerprints, and perform temperature prediction based on the temperature fingerprints to obtain the temperature of the target to be measured.

[0012] Furthermore, the passive temperature sensor includes an RFID tag, which is attached to or near the surface of the target to be measured, for sensing temperature changes of the target. The RFID tag receives a wireless radio frequency signal emitted by the temperature sensing unit and modulates the wireless radio frequency signal with ON-OFF backscatter by changing the chip impedance state. When the temperature of the target changes, the coupling effect between the RFID tag and the target changes, causing the antenna impedance and / or impedance mismatch coefficient of the RFID tag to change, and the backscatter response of the RFID tag at different frequencies to change.

[0013] Furthermore, the temperature sensing integrated machine includes: an RFID antenna, a radio frequency signal transmitting module, and a radio frequency signal receiving module;

[0014] The radio frequency signal transmitting module is used to generate and transmit multi-frequency wireless radio frequency signals;

[0015] The RFID antenna is used to transmit multi-frequency radio frequency signals to the passive temperature sensor and to receive the radio frequency signals backscattered by the passive temperature sensor.

[0016] The radio frequency signal receiving module is used to collect the in-phase and quadrature components of the backscattered radio frequency signal, obtain the received signal in complex form, and send the received signal to the data processing platform.

[0017] Furthermore, the multi-frequency radio frequency signal is an OFDM symbol, and the specific generation steps are as follows:

[0018] Generate length is pseudo-random sequence Each element has a 50% probability of being +1 or -1. For frequency domain subcarrier index;

[0019] For the pseudo-random sequence Perform an inverse fast Fourier transform shift (IFFT) to obtain the frequency domain OFDM sequence. , ;

[0020] For frequency domain OFDM sequences Perform inverse fast Fourier transform to obtain time-domain OFDM symbols. , ,in, OFDM symbol length, For frequency domain subcarrier index, For time-domain sampling point index, It is a natural constant. For imaginary units;

[0021] In time-domain OFDM symbols Add a length of to the end The cyclic suffix is ​​used to obtain the OFDM symbol with the cyclic suffix added, denoted as . The cyclic suffix is ​​formed by the prefix of the time-domain OFDM symbol. Each sampling point was copied. For the index of the cyclic suffix sampling point, This indicates the OFDM symbol after adding a cyclic suffix. One sampling point, This indicates the first digit in the OFDM symbol after adding a cyclic suffix. The complex value corresponding to each sampling point Represents the first in the original time-domain OFDM symbol The complex value corresponding to each sampling point;

[0022] According to the preset sampling rate and preset center frequency Repeated transmission of OFDM symbols with a cyclic suffix is ​​used to activate RFID tags and provide multiple orthogonal frequency channels.

[0023] Furthermore, the method by which the data processing platform obtains the temperature result data of the target to be measured is as follows:

[0024] (1) Acquire the received signals of RFID tags at different radio frequency signal frequencies ,in, , For the first Each radio frequency signal frequency, This represents the total number of radio frequency signal frequencies.

[0025] (2) Perform subcarrier synchronization on the received signals at different frequencies to obtain the synchronized received signals. ;

[0026] (3) From the synchronized received signal The ON state signal of the RFID tag is identified. and OFF status signal ;

[0027] (4) Eliminate random phase shift based on ON and OFF state signals to obtain the tag response signal after eliminating random phase shift. ;

[0028] (5) For any two different frequencies and The ratio of the tag response signals is calculated to obtain motion-independent temperature characteristics. ;

[0029] (6) Calculate the motion-independent temperature characteristics of all frequency pairs within the measurement range and construct a temperature fingerprint. ;

[0030] (7) Temperature fingerprint Input the temperature prediction model to obtain the temperature value of the target.

[0031] Furthermore, the specific steps for synchronizing the subcarriers in step (2) are as follows;

[0032] (21) Perform a Fast Fourier Transform on the received OFDM symbols to obtain the frequency domain received symbols. ,as follows:

[0033] ;

[0034] in, For the received time-domain OFDM symbol, the first The complex value corresponding to each sampling point The sampling point offset generated when the receiver determines the starting position of the OFDM symbol;

[0035] (22) Based on the cyclic suffix property, the frequency domain received symbol is represented as ,in, The frequency domain symbol when no sampling point shift occurs;

[0036] (23) Based on the known transmitter frequency domain OFDM sequence With frequency domain received symbols The phase relationship between them is used to calculate the index of different subcarriers. Phase offset on ,as follows:

[0037] ;

[0038] in, Indicates the angle used to calculate complex numbers;

[0039] (24) Regarding the phase offset With subcarrier index Perform linear fitting to obtain the sampling point offset. ;

[0040] (25) Based on the sampling point offset Adjusting the OFDM symbol start position yields the received signal after subcarrier synchronization. .

[0041] Furthermore, the specific steps for identifying the ON and OFF state signals of the RFID tag in step (3) are as follows:

[0042] (31) According to the RFID communication protocol, determine the time period after the EPC phase ends when the temperature sensing unit sends a continuous wave and the RFID tag stops backscattering modulation. ;

[0043] (32) Extraction OFDM symbols within the segment serve as OFF state reference signals. ;

[0044] (33) For any target OFDM symbol within the EPC phase, calculate its frequency. The amplitude below , and calculate The reference symbol for the OFF state within the segment at the frequency The amplitude below , ;

[0045] (34) Comparison and If the difference between the two is less than a preset threshold Then the target OFDM symbol is determined to be an OFF state signal. Otherwise, the target OFDM symbol is determined to be in the ON state. .

[0046] Furthermore, the specific steps for eliminating random phase shifts in step (4) are as follows:

[0047] Dividing the tag's backscattered component by the OFF state signal yields the tag response signal after eliminating random phase shift. ,as follows:

[0048] (1);

[0049] ;

[0050] ;

[0051] in, Excludes items related to radio frequency transceiver links , For RFID tag response related items, For terms related to the coupling path from the RFID tag to the temperature target, To leak path-related items, For items related to the radio frequency transceiver link, The temperature of the target object to be measured.

[0052] Furthermore, the specific steps for obtaining motion-independent temperature characteristics in step (5) are as follows:

[0053] (51) For any two different frequencies and The tag response signals after eliminating random phase shifts were obtained respectively. , ;

[0054] (52) Calculate the tag response signal With tag response signal The ratio of the frequencies yields the frequency-to-frequency ratio. Corresponding motion-independent temperature characteristics ,as follows:

[0055] (2);

[0056] (53) Substituting equation (1) into equation (2), we get the following:

[0057] ;

[0058] in, Represents frequency The following are items related to the leakage path. Represents frequency and temperature The following are the relevant items for RFID tag response. Represents frequency and temperature The coupling path related terms between the RFID tag and the target under test;

[0059] (54) Similar frequencies and The path-dependent terms below have similar responses. After eliminating the path-dependent terms, we obtain the following:

[0060] ;

[0061] in, Indicates temperature and frequency The impedance mismatch coefficient of the RFID tag. Indicates temperature and frequency Energy absorption parameters of the RFID tag; Indicates temperature and frequency The impedance mismatch coefficient of the RFID tag. Indicates temperature and frequency Energy absorption parameters of the RFID tag;

[0062] (55) Motion-independent temperature characteristics As frequency pair The corresponding motion-independent temperature characteristics.

[0063] Furthermore, the specific steps for constructing the temperature fingerprint in step (6) are as follows:

[0064] (61) For the frequency set Any two different frequencies and Combine them in pairs. ,in, , , This represents the total number of radio frequency signal frequencies.

[0065] (62) Calculate each frequency pair Corresponding motion-independent temperature characteristics ;

[0066] (63) Arrange all frequency pairs and their corresponding motion-independent temperature features in a preset order to obtain temperature fingerprints. as follows:

[0067] ;

[0068] Among them, temperature fingerprint Dimensions for:

[0069] ;

[0070] (64) The temperature fingerprint Reconstructed into a two-dimensional temperature fingerprint matrix , , This represents the operation of converting a one-dimensional vector into a two-dimensional matrix.

[0071] Furthermore, step (7) specifically includes

[0072] Two-dimensional temperature fingerprint matrix Input a temperature prediction model, which is a ThermoNet model, including an embedding network and a regression network;

[0073] Feature embedding network for temperature fingerprinting Mapped to latent feature vectors ,as follows:

[0074] ;

[0075] in, Indicates feature embedding network, These represent the parameters of the feature embedding network;

[0076] Regression networks are used to transform latent feature vectors Predicted temperature value mapped to the target ,as follows:

[0077] ;

[0078] in, Indicating a return to the internet, This represents the parameters of the regression network.

[0079] This invention also provides a robust temperature monitoring method based on RFID passive multi-frequency sensing. Based on the above system, the steps are as follows:

[0080] 1) Attach the passive temperature sensor to or near the surface of the target to be measured;

[0081] 2) Generate multi-frequency OFDM symbols and continuously transmit multi-frequency OFDM symbols to activate the passive temperature sensor;

[0082] 3) Real-time acquisition of the backscattered radio frequency signal from the passive temperature sensor to obtain the received signal in complex form;

[0083] 4) Perform EPC stage signal separation and subcarrier synchronization on the received signal to obtain synchronized received signals at multiple frequencies;

[0084] 5) Identify the ON and OFF status signals of RFID tags;

[0085] 6) Perform random phase offset elimination based on the ON and OFF state signals to obtain the tag response signal after eliminating random phase offset;

[0086] 7) Calculate the ratio of the tag response signals at any two different frequencies to obtain the motion-independent temperature characteristics;

[0087] 8) Combine all frequency pairs and their corresponding motion-independent temperature features into a temperature fingerprint;

[0088] 9) Input the temperature fingerprint into the temperature prediction model to obtain the temperature value of the target.

[0089] The beneficial effects of this invention are:

[0090] 1. This invention utilizes the characteristic that a single RFID tag has similar responses to equipment movement and environmental movement at similar frequencies, but different responses to temperature changes. By eliminating path-related interference through multi-frequency ratio calculation or differential calculation, the impact of equipment movement, target movement, and environmental multipath changes on RFID radio frequency signals is effectively reduced, thereby improving the stability of temperature monitoring in dynamic scenarios.

[0091] 2. This invention uses OFDM technology to provide multiple orthogonal frequency channels simultaneously. Compared with the traditional RFID reader's sequential frequency hopping acquisition method, it can obtain tag response signals at multiple frequencies in a shorter time, avoiding the problem of inconsistent acquisition environments of different frequency signals due to frequency switching delays in dynamic scenarios, and improving the real-time performance and reliability of multi-frequency temperature feature extraction.

[0092] 3. This invention establishes a physical layer signal model based on the ON and OFF state signals of RFID tags. By dividing the difference between the ON and OFF state signals by the OFF state signal, the random phase shift introduced by the instability of the radio frequency transceiver link is eliminated, making the obtained tag response signal more stable and improving the accuracy of temperature fingerprint extraction.

[0093] 4. This invention combines motion-independent temperature features corresponding to multiple frequency pairs into a temperature fingerprint, and uses a temperature prediction model to enhance the features and perform regression prediction on the temperature fingerprint, thereby improving the distinguishability of temperature fingerprints under similar temperatures and thus improving the accuracy of temperature prediction.

[0094] 5. This invention uses a single RFID tag to complete temperature sensing, without the need for battery power or additional deployment of reference tags or dedicated temperature sensors. It can be directly applied to scenarios such as cold chain logistics, warehouse management, conveyor belt object monitoring, and human health monitoring, and has the advantages of simple deployment, low cost, and suitability for large-scale applications. Attached Figure Description

[0095] Figure 1 This is a schematic diagram of the system architecture of the present invention.

[0096] Figure 2 This is a flowchart illustrating the principle of the method of the present invention. Detailed Implementation

[0097] To facilitate understanding by those skilled in the art, the present invention will be further described below with reference to embodiments and accompanying drawings. The content mentioned in the embodiments is not intended to limit the present invention.

[0098] Reference Figure 1 As shown, the present invention provides a motion-robust temperature monitoring system based on RFID passive multi-frequency sensing, comprising: a passive temperature sensor, an integrated temperature sensing unit, and a data processing platform.

[0099] A passive temperature sensor is used to receive the wireless radio frequency signal emitted by the temperature sensing all-in-one machine, and then transmit the received wireless radio frequency signal to the temperature sensing all-in-one machine after backscattering modulation.

[0100] Specifically, the passive temperature sensor includes an RFID tag, which is attached to or near the surface of the target to be measured, for sensing temperature changes of the target. The RFID tag receives a wireless radio frequency signal emitted by the temperature sensing unit and modulates the wireless radio frequency signal by changing the chip impedance state through ON-OFF backscatter modulation. When the temperature of the target changes, the coupling effect between the RFID tag and the target changes, causing the antenna impedance and / or impedance mismatch coefficient of the RFID tag to change, and the backscatter response of the RFID tag at different frequencies to change.

[0101] The temperature sensing all-in-one machine is used to generate multi-frequency wireless radio frequency signals to activate the passive temperature sensor, and to collect the backscattered wireless radio frequency signals of the passive temperature sensor in real time and send them to the data processing platform.

[0102] Specifically, the temperature sensing integrated machine includes: an RFID antenna, a radio frequency signal transmitting module, and a radio frequency signal receiving module;

[0103] The radio frequency signal transmitting module is used to generate and transmit multi-frequency wireless radio frequency signals;

[0104] The RFID antenna is used to transmit multi-frequency radio frequency signals to the passive temperature sensor and to receive the radio frequency signals backscattered by the passive temperature sensor.

[0105] The radio frequency signal receiving module is used to collect the in-phase and quadrature components of the backscattered radio frequency signal, obtain the received signal in complex form, and send the received signal to the data processing platform.

[0106] The multi-frequency radio frequency signal is an OFDM symbol, and the specific generation steps are as follows:

[0107] Generate length is pseudo-random sequence Each element has a 50% probability of being +1 or -1. For frequency domain subcarrier index;

[0108] For the pseudo-random sequence Perform an inverse fast Fourier transform shift (IFFT) to obtain the frequency domain OFDM sequence. , ;

[0109] For frequency domain OFDM sequences Perform inverse fast Fourier transform to obtain time-domain OFDM symbols. , ,in, OFDM symbol length, For frequency domain subcarrier index, For time-domain sampling point index, It is a natural constant. For imaginary units;

[0110] In time-domain OFDM symbols Add a length of to the end The cyclic suffix is ​​used to obtain the OFDM symbol with the cyclic suffix added, denoted as . The cyclic suffix is ​​formed by the prefix of the time-domain OFDM symbol. Each sampling point was copied. For the index of the cyclic suffix sampling point, This indicates the OFDM symbol after adding a cyclic suffix. One sampling point, This indicates the first digit in the OFDM symbol after adding a cyclic suffix. The complex value corresponding to each sampling point Represents the first in the original time-domain OFDM symbol The complex value corresponding to each sampling point;

[0111] According to the preset sampling rate and preset center frequency Repeated transmission of OFDM symbols with a cyclic suffix is ​​used to activate RFID tags and provide multiple orthogonal frequency channels.

[0112] The data processing platform is used to acquire the wireless radio frequency signal sent by the temperature sensing all-in-one machine in real time, perform subcarrier synchronization, random phase offset elimination and multi-frequency differential processing on the wireless radio frequency signal, extract motion-independent temperature fingerprint, and perform temperature prediction based on the temperature fingerprint to obtain the temperature of the target to be measured.

[0113] Specifically, the method by which the data processing platform obtains the temperature result data of the target to be measured is as follows:

[0114] (1) Acquire the received signals of RFID tags at different radio frequency signal frequencies ,in, , For the first Each radio frequency signal frequency, This represents the total number of radio frequency signal frequencies.

[0115] (2) Perform subcarrier synchronization on the received signals at different frequencies to obtain the synchronized received signals. ;

[0116] (3) From the synchronized received signal The ON state signal of the RFID tag is identified. and OFF status signal ;

[0117] (4) Eliminate random phase shift based on ON and OFF state signals to obtain the tag response signal after eliminating random phase shift. ;

[0118] (5) For any two different frequencies and The ratio of the tag response signals is calculated to obtain motion-independent temperature characteristics. ;

[0119] (6) Calculate the motion-independent temperature characteristics of all frequency pairs within the measurement range and construct a temperature fingerprint. ;

[0120] (7) Temperature fingerprint Input the temperature prediction model to obtain the temperature value of the target.

[0121] The specific steps for synchronizing the subcarriers in step (2) are as follows;

[0122] (21) Perform a Fast Fourier Transform on the received OFDM symbols to obtain the frequency domain received symbols. ,as follows:

[0123] ;

[0124] in, For the received time-domain OFDM symbol, the first The complex value corresponding to each sampling point The sampling point offset generated when the receiver determines the starting position of the OFDM symbol;

[0125] (22) Based on the cyclic suffix property, the frequency domain received symbol is represented as ,in, The frequency domain symbol when no sampling point shift occurs;

[0126] (23) Based on the known transmitter frequency domain OFDM sequence With frequency domain received symbols The phase relationship between them is used to calculate the index of different subcarriers. Phase offset on ,as follows:

[0127] ;

[0128] in, Indicates the angle used to calculate complex numbers;

[0129] (24) Regarding the phase offset With subcarrier index Perform linear fitting to obtain the sampling point offset. ;

[0130] (25) Based on the sampling point offset Adjusting the OFDM symbol start position yields the received signal after subcarrier synchronization. .

[0131] The specific steps for identifying the ON and OFF status signals of the RFID tag in step (3) are as follows:

[0132] (31) According to the RFID communication protocol, determine the time period after the EPC phase ends when the temperature sensing unit sends a continuous wave and the RFID tag stops backscattering modulation. ;

[0133] (32) Extraction OFDM symbols within the segment serve as OFF state reference signals. ;

[0134] (33) For any target OFDM symbol within the EPC phase, calculate its frequency. The amplitude below , and calculate The reference symbol for the OFF state within the segment at the frequency The amplitude below , ;

[0135] (34) Comparison and If the difference between the two is less than a preset threshold Then the target OFDM symbol is determined to be an OFF state signal. Otherwise, the target OFDM symbol is determined to be in the ON state. .

[0136] The specific steps for eliminating random phase shifts in step (4) are as follows:

[0137] Dividing the tag's backscattered component by the OFF state signal yields the tag response signal after eliminating random phase shift. ,as follows:

[0138] (1);

[0139] ;

[0140] ;

[0141] in, Excludes items related to radio frequency transceiver links , For RFID tag response related items, For terms related to the coupling path from the RFID tag to the temperature target, To leak path-related items, For items related to the radio frequency transceiver link, The temperature of the target object to be measured.

[0142] The specific steps for obtaining motion-independent temperature characteristics in step (5) are as follows:

[0143] (51) For any two different frequencies and The tag response signals after eliminating random phase shifts were obtained respectively. , ;

[0144] (52) Calculate the tag response signal With tag response signal The ratio of the frequencies yields the frequency-to-frequency ratio. Corresponding motion-independent temperature characteristics ,as follows:

[0145] (2);

[0146] (53) Substituting equation (1) into equation (2), we get the following:

[0147] ;

[0148] in, Represents frequency The following are items related to the leakage path. Represents frequency and temperature The following are the relevant items for RFID tag response. Represents frequency and temperature The coupling path related terms between the RFID tag and the target under test;

[0149] (54) Similar frequencies and The path-dependent terms below have similar responses. After eliminating the path-dependent terms, we obtain the following:

[0150] ;

[0151] in, Indicates temperature and frequency The impedance mismatch coefficient of the RFID tag. Indicates temperature and frequency Energy absorption parameters of the RFID tag; Indicates temperature and frequency The impedance mismatch coefficient of the RFID tag. Indicates temperature and frequency Energy absorption parameters of the RFID tag;

[0152] (55) Motion-independent temperature characteristics As frequency pair The corresponding motion-independent temperature characteristics.

[0153] The specific steps for constructing the temperature fingerprint in step (6) are as follows:

[0154] (61) For the frequency set Any two different frequencies and Combine them in pairs. ,in, , , This represents the total number of radio frequency signal frequencies.

[0155] (62) Calculate each frequency pair Corresponding motion-independent temperature characteristics ;

[0156] (63) Arrange all frequency pairs and their corresponding motion-independent temperature features in a preset order to obtain temperature fingerprints. as follows:

[0157] ;

[0158] Among them, temperature fingerprint Dimensions for:

[0159] ;

[0160] (64) The temperature fingerprint Reconstructed into a two-dimensional temperature fingerprint matrix , , This represents the operation of converting a one-dimensional vector into a two-dimensional matrix.

[0161] Specifically, step (7) includes

[0162] Two-dimensional temperature fingerprint matrix Input a temperature prediction model, which is a ThermoNet model, including an embedding network and a regression network;

[0163] Feature embedding network for temperature fingerprinting Mapped to latent feature vectors ,as follows:

[0164] ;

[0165] in, Indicates feature embedding network, These represent the parameters of the feature embedding network;

[0166] Regression networks are used to transform latent feature vectors Predicted temperature value mapped to the target ,as follows:

[0167] ;

[0168] in, Indicating a return to the internet, This represents the parameters of the regression network.

[0169] Reference Figure 2 As shown, the present invention also provides a robust temperature monitoring method based on RFID passive multi-frequency sensing. Based on the above system, the steps are as follows:

[0170] 1) Attach the passive temperature sensor to or near the surface of the target to be measured;

[0171] 2) Generate multi-frequency OFDM symbols and continuously transmit multi-frequency OFDM symbols to activate the passive temperature sensor;

[0172] 3) Real-time acquisition of the backscattered radio frequency signal from the passive temperature sensor to obtain the received signal in complex form;

[0173] 4) Perform EPC stage signal separation and subcarrier synchronization on the received signal to obtain synchronized received signals at multiple frequencies;

[0174] 5) Identify the ON and OFF status signals of RFID tags;

[0175] 6) Perform random phase offset elimination based on the ON and OFF state signals to obtain the tag response signal after eliminating random phase offset;

[0176] 7) Calculate the ratio of the tag response signals at any two different frequencies to obtain the motion-independent temperature characteristics;

[0177] 8) Combine all frequency pairs and their corresponding motion-independent temperature features into a temperature fingerprint;

[0178] 9) Input the temperature fingerprint into the temperature prediction model to obtain the temperature value of the target.

[0179] This invention has many specific applications. The above description is only a preferred embodiment of this invention. It should be noted that for those skilled in the art, several improvements can be made without departing from the principle of this invention, and these improvements should also be considered within the scope of protection of this invention.

Claims

1. A motion-robust temperature monitoring system based on RFID passive multi-frequency sensing, characterized in that, include: Passive temperature sensor, integrated temperature sensing unit and data processing platform; A passive temperature sensor is used to receive the wireless radio frequency signal emitted by the temperature sensing all-in-one machine, and then transmit the received wireless radio frequency signal to the temperature sensing all-in-one machine after backscattering modulation. The temperature sensing all-in-one machine is used to generate multi-frequency wireless radio frequency signals to activate the passive temperature sensor, and to collect the backscattered wireless radio frequency signals of the passive temperature sensor in real time and send them to the data processing platform. The data processing platform is used to acquire the wireless radio frequency signal sent by the temperature sensing all-in-one machine in real time, perform subcarrier synchronization, random phase offset elimination and multi-frequency differential processing on the wireless radio frequency signal, extract motion-independent temperature fingerprints, and perform temperature prediction based on the temperature fingerprints to obtain the temperature of the target to be measured.

2. The motion-robust temperature monitoring system based on RFID passive multi-frequency sensing according to claim 1, characterized in that, The passive temperature sensor includes an RFID tag attached to or near the surface of the target to be measured, used to sense temperature changes in the target. The RFID tag receives a radio frequency signal emitted by the temperature sensing unit and modulates the radio frequency signal with ON-OFF backscatter by changing the chip impedance state. When the temperature of the target changes, the coupling effect between the RFID tag and the target changes, causing the antenna impedance and / or impedance mismatch coefficient of the RFID tag to change, and the backscatter response of the RFID tag at different frequencies to change.

3. The motion-robust temperature monitoring system based on RFID passive multi-frequency sensing according to claim 1, characterized in that, The temperature sensing integrated machine includes: an RFID antenna, a radio frequency signal transmitting module, and a radio frequency signal receiving module; The radio frequency signal transmitting module is used to generate and transmit multi-frequency wireless radio frequency signals; The RFID antenna is used to transmit multi-frequency radio frequency signals to the passive temperature sensor and to receive the radio frequency signals backscattered by the passive temperature sensor. The radio frequency signal receiving module is used to collect the in-phase and quadrature components of the backscattered radio frequency signal, obtain the received signal in complex form, and send the received signal to the data processing platform.

4. The motion-robust temperature monitoring system based on RFID passive multi-frequency sensing according to claim 3, characterized in that, The multi-frequency radio frequency signal is an OFDM symbol, and the specific generation steps are as follows: Generate length is pseudo-random sequence Each element has a 50% probability of being +1 or -1. For frequency domain subcarrier index; For the pseudo-random sequence Perform an inverse fast Fourier transform shift to obtain the frequency domain OFDM sequence. , ; For frequency domain OFDM sequences Perform inverse fast Fourier transform to obtain time-domain OFDM symbols. ; ； in, OFDM symbol length, For frequency domain subcarrier index, For time-domain sampling point index, It is a natural constant. For imaginary units; In time-domain OFDM symbols Add a length of to the end The cyclic suffix is ​​used to obtain the OFDM symbol with the cyclic suffix added, denoted as . The cyclic suffix is ​​formed by the prefix of the time-domain OFDM symbol. Each sampling point was copied. For the index of the cyclic suffix sampling point, This indicates the OFDM symbol after adding a cyclic suffix. One sampling point, This indicates the first digit in the OFDM symbol after adding a cyclic suffix. The complex value corresponding to each sampling point Represents the first in the original time-domain OFDM symbol The complex value corresponding to each sampling point; According to the preset sampling rate and preset center frequency Repeated transmission of OFDM symbols with a cyclic suffix is ​​used to activate RFID tags and provide multiple orthogonal frequency channels.

5. The motion-robust temperature monitoring system based on RFID passive multi-frequency sensing according to claim 1, characterized in that, The method by which the data processing platform obtains the temperature result data of the target to be measured is as follows: (1) Acquire the received signals of RFID tags at different radio frequency signal frequencies ,in, , For the first Each radio frequency signal frequency, This represents the total number of radio frequency signal frequencies. (2) Perform subcarrier synchronization on the received signals at different frequencies to obtain the synchronized received signals. ; (3) From the synchronized received signal The ON state signal of the RFID tag is identified. and OFF status signal ; (4) Eliminate random phase shift based on ON and OFF state signals to obtain the tag response signal after eliminating random phase shift. ; (5) For any two different frequencies and The ratio of the tag response signals is calculated to obtain motion-independent temperature characteristics. ; (6) Calculate the motion-independent temperature characteristics of all frequency pairs within the measurement range and construct a temperature fingerprint. ; (7) Temperature fingerprint Input the temperature prediction model to obtain the temperature value of the target.

6. The motion-robust temperature monitoring system based on RFID passive multi-frequency sensing according to claim 5, characterized in that, The specific steps for synchronizing the subcarriers in step (2) are as follows; (21) Perform a Fast Fourier Transform on the received OFDM symbols to obtain the frequency domain received symbols. ,as follows: ； in, For the received time-domain OFDM symbol, the first The complex value corresponding to each sampling point The sampling point offset generated when the receiver determines the starting position of the OFDM symbol; (22) Based on the cyclic suffix property, the frequency domain received symbol is represented as ,in, The frequency domain symbol when no sampling point shift occurs; (23) Based on the known transmitter frequency domain OFDM sequence With frequency domain received symbols The phase relationship between them is used to calculate the index of different subcarriers. Phase offset on ,as follows: ； in, Indicates the angle used to calculate complex numbers; (24) Regarding the phase offset With subcarrier index Perform linear fitting to obtain the sampling point offset. ; (25) Based on the sampling point offset Adjusting the OFDM symbol start position yields the received signal after subcarrier synchronization. .

7. The motion-robust temperature monitoring system based on RFID passive multi-frequency sensing according to claim 6, characterized in that, The specific steps for identifying the ON and OFF status signals of the RFID tag in step (3) are as follows: (31) According to the RFID communication protocol, determine the time period after the EPC phase ends when the temperature sensing unit sends a continuous wave and the RFID tag stops backscattering modulation. ; (32) Extraction OFDM symbols within the segment serve as OFF state reference signals. ; (33) For any target OFDM symbol within the EPC phase, calculate its frequency. The amplitude below , and calculate The reference symbol for the OFF state within the segment at the frequency The amplitude below , ; (34) Comparison and If the difference between the two is less than a preset threshold Then the target OFDM symbol is determined to be an OFF state signal. Otherwise, the target OFDM symbol is determined to be in the ON state. .

8. The motion-robust temperature monitoring system based on RFID passive multi-frequency sensing according to claim 7, characterized in that, The specific steps for eliminating random phase shifts in step (4) are as follows: Dividing the tag's backscattered component by the OFF state signal yields the tag response signal after eliminating random phase shifts. ,as follows: （1）； ； ； in, Excludes items related to radio frequency transceiver links , For RFID tag response related items, For terms related to the coupling path from the RFID tag to the temperature target, To leak path-related items, For items related to the radio frequency transceiver link, The temperature of the target object to be measured.

9. The motion-robust temperature monitoring system based on RFID passive multi-frequency sensing according to claim 8, characterized in that, The specific steps for obtaining motion-independent temperature characteristics in step (5) are as follows: (51) For any two different frequencies and The tag response signals after eliminating random phase shifts were obtained respectively. , ; (52) Calculate the tag response signal With tag response signal The ratio of the frequencies yields the frequency-to-frequency ratio. Corresponding motion-independent temperature characteristics ,as follows: （2）； (53) Substituting equation (1) into equation (2), we get the following: ； in, Represents frequency The following are items related to the leakage path. Represents frequency and temperature The following are the relevant items for RFID tag response. Represents frequency and temperature The coupling path related terms between the RFID tag and the target under test; (54) Similar frequencies and The path-dependent terms below have similar responses. After eliminating the path-dependent terms, we obtain the following: ； in, Indicates temperature and frequency The impedance mismatch coefficient of the RFID tag. Indicates temperature and frequency Energy absorption parameters of the RFID tag; Indicates temperature and frequency The impedance mismatch coefficient of the RFID tag. Indicates temperature and frequency Energy absorption parameters of the RFID tag; (55) Motion-independent temperature characteristics As frequency pair The corresponding motion-independent temperature characteristics.

10. A robust temperature monitoring method based on RFID passive multi-frequency sensing, based on the system described in any one of claims 1-9, characterized in that, The steps are as follows: 1) Attach the passive temperature sensor to or near the surface of the target to be measured; 2) Generate multi-frequency OFDM symbols and continuously transmit multi-frequency OFDM symbols to activate the passive temperature sensor; 3) Real-time acquisition of the backscattered radio frequency signal from the passive temperature sensor to obtain the received signal in complex form; 4) Perform EPC stage signal separation and subcarrier synchronization on the received signal to obtain synchronized received signals at multiple frequencies; 5) Identify the ON and OFF status signals of RFID tags; 6) Perform random phase offset elimination based on the ON and OFF state signals to obtain the tag response signal after eliminating random phase offset; 7) Calculate the ratio of the tag response signals at any two different frequencies to obtain the motion-independent temperature characteristics; 8) Combine all frequency pairs and their corresponding motion-independent temperature features into a temperature fingerprint; 9) Input the temperature fingerprint into the temperature prediction model to obtain the temperature value of the target.

Images