A temperature-pressure measurement method based on gas density driven sensing

Abstract

The application discloses a temperature-pressure measurement method based on gas density driving sensing, and belongs to the technical field of flexible electronics and intelligent sensing.The application is based on the influence of temperature and pressure on the gas density in a sealed air bag, that is, the increase of external temperature load leads to the expansion of gas and the decrease of gas density, on the contrary, the decrease of temperature leads to the contraction of gas and the increase of gas density; the increase of external pressure compresses the air bag and increases the density. Therefore, the influence of temperature and pressure on the gas density presents differentiated adjustment characteristics in physical action, that is, the contribution of temperature change and pressure change to the gas density has distinguishable physical characteristics. The application utilizes the distinguishable physical effects of temperature and pressure on the gas density in the sealed air bag, combines machine learning, frequency domain decoupling and other methods, and can efficiently and robustly realize the independent and synchronous measurement of temperature and pressure.

Description

Technical Field

[0001] This invention belongs to the field of flexible electronics and intelligent sensing technology, and specifically relates to a temperature-pressure measurement method based on gas density-driven sensing. Background Technology

[0002] In cutting-edge fields such as flexible electronics, wearable devices, and soft robotics, the ability to simultaneously and independently sense external temperature and mechanical pressure loads is crucial for achieving advanced human-machine interaction and intelligent environmental perception. This type of bimodal information is essential for real-time and accurate acquisition of external composite temperature and pressure load signals or for endowing robots with fine tactile sensation.

[0003] Currently, most mainstream flexible sensors for temperature and pressure detection employ principles such as piezoresistive, capacitive, or piezoelectric. However, these sensing mechanisms are typically designed to detect a single physical quantity (such as pressure or temperature alone). In practical applications, temperature and pressure often act on the sensing unit simultaneously, causing their signals to interfere with each other, making it difficult to accurately distinguish their respective contributions—a severe "temperature-pressure coupling" effect.

[0004] To address this issue, existing technologies generally employ a multi-sensor array layout, where independent temperature and pressure sensing units are placed within a single device. However, this approach not only significantly increases the structural complexity and fabrication cost of the sensor but also easily leads to crosstalk and spatial interference due to the physical proximity of the two types of functional units, potentially reducing overall detection accuracy. Furthermore, some studies have attempted post-decoupling through complex signal compensation algorithms, but such methods typically rely on large amounts of calibration data and complex computational models, resulting in slow response times and insufficient robustness in dynamic scenarios with rapid and synchronous changes in temperature and pressure.

[0005] Therefore, there is an urgent need for a new measurement method that has a clear physical mechanism for responding to temperature and pressure and can effectively adapt to complex working conditions where the two dynamically co-varie. Summary of the Invention

[0006] In view of this, in order to solve the problems existing in the prior art, the purpose of this invention is to provide a temperature-pressure measurement method based on gas density driven sensing, so as to utilize the distinguishable physical effects of temperature and pressure on the gas density in the sealed airbag, and combine machine learning, frequency domain decoupling and other methods to achieve independent and synchronous measurement of temperature and pressure in an efficient and robust manner.

[0007] The technical method of this invention is a temperature-pressure measurement method based on gas density-driven sensing, characterized by the following steps:

[0008] Step 1: Collect the gas density response signal inside the sealed airbag. Using gas density as the detection target and the change in gas density as the sensing response signal, the observation vector at the current moment is formed.

[0009] The specific steps for acquiring multidimensional gas density response signals are as follows:

[0010] The sensing airbag unit is placed in the environment to be tested, and the gas density response signal of each sealed airbag is collected in real time to form a gas density response vector. The gas density response vector is composed of the gas density response values ​​of multiple airbags, where the gas density of the k-th thermo-baric response of the N-th airbag is expressed as... The gas density of the Nth airbag is represented by the subscript r, which characterizes the airbag's response to external temperature and pressure loads, and the superscript (k) characterizes the kth temperature and pressure response measurement.

[0011] Step 2: Establish a nonlinear mapping model to map the gas density response signal to the corresponding applied temperature load and external pressure load;

[0012] Step 3: Perform synchronous decoupling calculations to obtain independent temperature and pressure values;

[0013] Based on the nonlinear mapping model of temperature-pressure-gas density constructed in step 2, and considering the dynamic characteristics of the temperature and pressure changes in the environment under test (static or dynamic, slow or fast), an appropriate decoupling calculation mode is selected to decouple the density response signal, thereby achieving independent and accurate solutions for the external temperature load T and pressure load F. The three modes are as follows:

[0014] Mode 1: Data-driven decoupling mode, adaptable to a wide range of complex dynamic temperature and pressure conditions, when there are no explicit physical model simplification conditions;

[0015] Mode 2: Frequency domain decoupling mode, adapted to dynamic temperature and pressure conditions, and different dynamic characteristic scenarios with slow temperature changes and rapid pressure changes;

[0016] Mode 3: Physical feature modeling decoupling mode, adapted to static / slow temperature and pressure change conditions, when high-precision decoupling is achieved by relying on physical mechanisms;

[0017] Furthermore, the specific method for step 2 is as follows:

[0018] Step 2.1: Determine the basic inputs for modeling;

[0019] The gas density response vectors of the N sealed airbags collected in step 1 The core observation input is included; at the same time, the inherent parameters of each airbag are also included: geometric dimensions D, shape factor S, and filling gas type parameter G. Different gas types correspond to different gas constants, thermal expansion coefficients and other characteristic parameters, which are used as fixed parameter inputs for modeling.

[0020] Step 2.2: Constructing the single airbag density response function: Based on the mechanism of temperature and pressure on gas density, a nonlinear function for the gas density response of a single airbag is established. The gas density response of the i-th airbag satisfies: in, Let be the gas density of the temperature-pressure response measured in the k-th time for the i-th airbag, where T is the external temperature load and F is the external pressure load. The nonlinear mapping function for the i-th airbag is obtained by combining the gas state equation and the mechanical deformation characteristics of the airbag's flexible wall.

[0021] Step 2.3: Constructing the nonlinear equation system for the array airbags: Based on the single airbag response functions of N heterogeneous airbags, integrate them to form a nonlinear equation system containing N equations, realizing the correlation between multi-dimensional density signals and temperature and pressure loads. The equation system is in the following form:

[0022] ;

[0023] Step 2.4: Achieve inverse solution of the mapping model: To ensure that temperature T and pressure F can be uniquely solved from the density response vector, the above nonlinear mapping relationship is optimized for effectiveness:

[0024] If the application scenario involves localized, small-scale temperature and pressure changes, then the nonlinear function... Within the working interval, a first-order Taylor expansion approximation is performed to complete the local linearization process, transforming the nonlinear equation system into a directly solvable linear equation system. Y is the density response vector, and X is... Temperature-pressure vector, A is the linearization coefficient matrix;

[0025] If the application scenario involves a wide range of temperature and pressure variations, multiple sets of temperature and pressure label-density response vector samples are obtained through system calibration experiments. Based on the samples, the nonlinear mapping relationship is fitted and corrected to establish a globally reversible mapping model, ensuring the uniqueness and accuracy of the inverse solution of the model throughout the entire working range.

[0026] Furthermore, the specific method for Mode 1 in step 3 is as follows:

[0027] Using the nonlinear mapping model from step 2 as a framework, and training the data model with calibration data, end-to-end decoupled calculation from density signal to temperature and pressure values ​​is directly achieved. Specific steps are as follows:

[0028] (1) Calibration dataset construction: The sensor airbag array is placed in a controllable temperature and pressure calibration environment, and multiple sets of discrete temperature and pressure points are selected in the entire working range. The density response vector of the airbag array was collected at each set of temperature and pressure points to construct a calibration dataset. ,in Let be the density response vector under the j-th temperature and pressure group, and m be the number of calibration samples. The number of samples must meet the requirements for suppressing overfitting during model training.

[0029] (2) Data model training: Select a machine learning / deep learning model as the decoupling model and calibrate the density response vector in the dataset. For input, the corresponding temperature ,pressure To output labels, the model is trained and validated. The model structure is adjusted through hyperparameter optimization and cross-validation to ensure that the error between the model's predicted values ​​and the actual temperature and pressure values ​​meets the measurement accuracy requirements, thus obtaining the trained data-driven decoupled model.

[0030] (3) Real-time decoupling calculation: Input the real-time gas density response vector collected in step 1 into the trained data-driven decoupling model. The model directly outputs the decoupled independent temperature value T and pressure value F to complete a temperature and pressure measurement.

[0031] Furthermore, the specific method for mode two in step 3 is as follows:

[0032] By utilizing the characteristic differences of temperature and pressure changes in the frequency domain, the mapping model in step 2 is reconstructed in the frequency domain. Decoupling of the temperature and pressure signals is achieved through spectral separation. Specific steps are as follows:

[0033] (1) Frequency domain feature calibration: In a controllable dynamic temperature and pressure environment, pure temperature dynamic excitation (constant pressure) and pure pressure dynamic excitation (constant temperature) are applied respectively. The density response time domain signals under the two excitations are collected. The time domain signals are converted into frequency domain signals by fast Fourier transform. The temperature characteristic frequency band and pressure characteristic frequency band are calibrated. At the same time, the quantitative mapping relationship between the density signal amplitude / phase and the temperature and pressure load in the two frequency bands is obtained by fitting. The low frequency band is (0.1Hz, 1Hz], and the high frequency band is (1-100Hz).

[0034] (2) Real-time signal spectrum conversion: The real-time multidimensional gas density response time domain signal collected in step 1 is denoised and smoothed. Then, Fourier transform is performed on the density response signal of each airbag to convert the time domain signal into the frequency domain signal and obtain the signal amplitude and phase at each frequency point.

[0035] (3) Frequency band signal separation: Based on the calibrated temperature and pressure characteristic frequency bands, a bandpass filtering algorithm is used to filter the frequency domain signal and separate the signal components within the temperature characteristic frequency band. and signal components within the pressure characteristic frequency band This eliminates the frequency domain superposition interference between the two.

[0036] (4) Temperature and pressure value inversion calculation: the separated temperature frequency band signal components Pressure frequency band signal components The independent temperature value T and pressure value F are obtained by substituting the corresponding frequency domain quantization mapping relationship into the inversion algorithm. The decoupling results of the multidimensional airbag are fused and averaged to improve the measurement accuracy.

[0037] Furthermore, the specific method for mode two in step 3 is as follows:

[0038] Based on the nonlinear equations from step 2, a high-dimensional eigenvector is constructed using the differentiated response characteristics of heterogeneous airbags to achieve analytical decoupled calculation of temperature and pressure loads. Specific steps include:

[0039] (1) Construction of high-dimensional feature vector: Based on the gas density response vectors of N heterogeneous airbags collected in step 1, and combined with the inherent parameter differences of each airbag, feature parameters such as the ratio of density response values ​​of different airbags in steady-state response and density change rate are extracted to construct a high-dimensional decoupled feature vector. , Let be the real-time density change rate of the i-th airbag;

[0040] (2) Global mapping relationship fitting: In a controllable temperature and pressure environment, high-dimensional feature vectors are collected at multiple temperature and pressure points throughout the entire working range. High-dimensional feature vectors are obtained by fitting them using a nonlinear fitting algorithm. With temperature and pressure load Global nonlinear analytical mapping relationship: This relationship is the characteristic simplified form of the nonlinear equations in step 2, which retains the core physical relationship between temperature, pressure and density.

[0041] (3) Analytical decoupling calculation: The real-time gas density response vector is transformed into the corresponding high-dimensional feature vector. Substitute the obtained analytical mapping relationship into the fitting. The analytical equations are solved by numerical iterative algorithms to obtain unique temperature values ​​T and pressure values ​​F, thus completing the decoupled calculation. If the nonlinear equations have been locally linearized in step 2, the temperature and pressure values ​​can be solved analytically directly by matrix inversion, improving computational efficiency.

[0042] This invention utilizes the thermo-mechanical coupling effect to achieve signal differentiation: It is based on the influence of temperature and pressure on the gas density within a sealed airbag—increased external temperature load leads to gas expansion and decreased gas density, while decreased temperature causes gas contraction and increased gas density; increased external pressure compresses the airbag, increasing density. Therefore, the effects of temperature and pressure on gas density exhibit differentiated regulatory characteristics in terms of physical action; that is, the contributions of temperature and pressure changes to gas density have distinguishable physical characteristics. This sensing mechanism based on changes in gas state provides a feasible technical approach for decoupling temperature and pressure signals, differing from traditional measurement methods that rely on changes in the electrical parameters of a single material (such as resistance and capacitance).

[0043] Supports simultaneous independent inversion of temperature and pressure: By constructing a gasbag response system with size or gas differences, differentiated outputs to temperature and pressure excitations are generated. Combined with modeling or algorithmic methods, independent temperature and pressure values ​​can be directly calculated under complex operating conditions where both temperature and pressure change simultaneously, without relying on the assumption that a certain parameter is constant. Crucially, the acquired temperature and pressure information strictly originates from the same spatial location and the same point in time, fundamentally different from schemes that integrate independent temperature / pressure sensing units within a single sensing device—the latter are prone to spatial inconsistencies or temporal hysteresis errors due to the physical spacing between sensing elements. This method effectively avoids such error sources, significantly improving measurement accuracy, reliability, and applicability in complex dynamic environments.

[0044] Fast response speed: The gas state responds quickly to external thermo-mechanical stimuli. Combined with an efficient signal processing flow, the system as a whole has fast response characteristics, making it particularly suitable for monitoring scenarios that require real-time capture of dynamic temperature and pressure changes.

[0045] The decoupling results are stable and reliable: Based on the physically distinguishable response mechanism, the decoupling process is not easily affected by cross-sensitivity interference. Even under temperature and pressure covariance conditions, it can still maintain good measurement consistency and repeatability, and has high robustness. Attached Figure Description

[0046] Figure 1 The diagram below shows the computational flow of a temperature-pressure measurement method based on gas density-driven sensing provided by this invention. Detailed Implementation

[0047] The technical solution adopted in this invention is: a temperature-pressure measurement method based on gas density-driven sensing, comprising the following steps:

[0048] 1. Acquire single / multidimensional gas density response signals

[0049] 2. Establish a nonlinear mapping model

[0050] 3. Perform synchronous decoupling computation

[0051] (1) Data-driven decoupling mode

[0052] (2) Frequency domain decoupling mode

[0053] (3) Decoupling mode of physical feature modeling

[0054] 4. Output decoupling results: Output the calculated independent temperature value T and pressure value F to complete a highly robust synchronous temperature and pressure measurement.

[0055] The relationship between temperature load and pressure load and gas density change:

[0056] Temperature response mechanism: When external temperature load As the temperature rises, the thermal motion of gas molecules intensifies, causing the flexible gasbag walls to expand outwards, resulting in an increase in gas volume. Increase, resulting in gas density Reduce; when external temperature load As the temperature drops, the thermal motion of gas molecules slows down, and the flexible airbag walls are compressed, leading to... Decrease, making Increase.

[0057] Pressure response mechanism: When the external pressure load F increases, the airbag is compressed. Decrease Enlargement, leading to Increase.

[0058] In summary, when using array airbags, the gas density response of the i-th airbag can be modeled as follows: .

[0059] The gas density of each airbag can be expressed as a function of temperature T, pressure F, its own size parameter D, shape S, gas type parameter G, and other relevant parameters.

[0060] An array of heterogeneous airbags can generate a system of nonlinear equations containing N equations:

[0061]

[0062] The effects of temperature and pressure on gas density differ significantly in terms of physical action; that is, the contributions of temperature changes and pressure changes to density have distinguishable physical characteristics.

[0063] Based on the dynamic characteristics of the application scenario, select any of the following decoupling modes to obtain the independent temperature T and pressure F values ​​at the current moment. (1) Data-driven decoupling mode (based on machine learning or deep learning); (2) Frequency domain decoupling mode; (3) Physical feature modeling decoupling mode (based on heterogeneous response to construct an analytical model). Output the calculated independent temperature value T and pressure value F to complete a highly robust synchronous temperature and pressure measurement.

[0064] Acquisition of single / multidimensional gas density response signals: A flexible temperature and pressure sensor is placed in the environment to be measured, and the gas density signal of one or more sealed air bladders inside is monitored in real time. The gas density signal is the result of the combined effect of gas pressure and volume inside the airbag, and is also affected by the external temperature load T and mechanical pressure load F.

[0065] The basic principle of a temperature-pressure measurement method based on gas density-driven sensing is as follows: This invention utilizes the response characteristics that increase in temperature causes gas expansion and decrease in gas density within an airbag, while increase in external pressure causes airbag compression and increase in gas density, thus forming distinguishable physical characteristics. By setting airbags of different sizes or filled with different gases, they can produce differentiated responses to the same temperature and pressure changes; by combining the response information of multiple airbags, independent temperature and pressure values ​​can be accurately retrieved even when temperature and pressure change simultaneously.

[0066] Constructing a nonlinear mapping relationship: Based on the physical mechanism, the gas density response of each airbag can be expressed as a nonlinear function of temperature T, pressure F, airbag size parameter D, shape S, filling gas type parameter G, and other related parameters. By setting up differentiated airbags (e.g., different in size or type of filling gas), they can produce distinguishable response characteristics to the same temperature and pressure excitation.

[0067] Perform synchronous decoupling calculations: Considering the dynamic characteristics of temperature and pressure changes in real-world applications, select one of the following modes for processing to obtain the independent temperature T and pressure F values ​​at the current moment:

[0068] a) Data-driven decoupling mode: During the calibration phase, the gas density response vector of the airbag array is collected under known temperature and pressure conditions, and this vector is used to train a machine learning or deep learning model; during actual use, the measured response is input into the model, and the decoupled T and F are directly output.

[0069] b) Frequency domain decoupling mode: Perform spectral analysis on the time-domain gas density signal, utilize the difference in dynamic response between temperature change (low frequency dominant) and pressure change (high frequency dominant), separate the mixed signal components in the frequency domain, and invert T and F by combining the pre-calibrated frequency characteristics;

[0070] c) Decoupling mode of physical feature modeling: Based on the difference in steady-state response amplitude of different airbags due to differences in size or gas, feature vectors are constructed, and a nonlinear mapping relationship between them and T and F is established through system calibration for real-time inversion.

[0071] Output decoupling results: Output the independent temperature value T and pressure value F obtained by the above decoupling calculation to complete a highly robust synchronous temperature and pressure measurement.

[0072] Through steps 1–4 above, this invention fully utilizes the opposite physical effects of temperature and pressure on the gas density within the airbag. Combined with the differentiated response of heterogeneous airbags, it can effectively achieve independent and synchronous sensing of temperature and pressure even under complex operating conditions where both are changing rapidly. This method is clear in principle, flexible in implementation, and exhibits good stability and environmental adaptability.

Claims

1. A temperature-pressure measurement method based on gas density driven sensing, characterized in that, Includes the following steps: Step 1: Collect the gas density response signal inside the sealed airbag. Using gas density as the detection target and the change in gas density as the sensing response signal, the observation vector at the current moment is formed. The specific steps for acquiring multidimensional gas density response signals are as follows: The sensing airbag unit is placed in the environment to be tested, and the gas density response signal of each sealed airbag is collected in real time to form a gas density response vector. The gas density response vector is composed of the gas density response values ​​of multiple airbags, where the gas density of the k-th thermo-baric response of the N-th airbag is expressed as... The gas density of the Nth airbag is represented by the subscript r, which characterizes the airbag's response to external temperature and pressure loads, and the superscript (k) characterizes the kth temperature and pressure response measurement. Step 2: Establish a nonlinear mapping model to map the gas density response signal to the corresponding applied temperature load and external pressure load; Step 3: Perform synchronous decoupling calculations to obtain independent temperature and pressure values; Based on the nonlinear mapping model of temperature-pressure-gas density constructed in step 2, and considering the dynamic characteristics of the temperature and pressure changes in the environment under test (static or dynamic, slow or fast), an appropriate decoupling calculation mode is selected to decouple the density response signal, thereby achieving independent and accurate solutions for the external temperature load T and pressure load F. The three modes are as follows: Mode 1: Data-driven decoupling mode, adaptable to a wide range of complex dynamic temperature and pressure conditions, when there are no explicit physical model simplification conditions; Mode 2: Frequency domain decoupling mode, adapted to dynamic temperature and pressure conditions, and different dynamic characteristic scenarios with slow temperature changes and rapid pressure changes; Mode 3: Physical feature modeling decoupling mode, adapted to static / slow temperature and pressure change conditions, requires high-precision decoupling based on physical mechanisms.

2. A temperature-pressure measurement method based on gas density drive sensing as claimed in claim 1, wherein, The specific method for step 2 is as follows: Step 2.1: Determine the basic inputs for modeling; The gas density response vectors of the N sealed airbags collected in step 1 The core observation input is included; at the same time, the inherent parameters of each airbag are also included: geometric dimensions D, shape factor S, and filling gas type parameter G. Different gas types correspond to different gas constants, thermal expansion coefficients and other characteristic parameters, which are used as fixed parameter inputs for modeling. Step 2.2: Constructing the single airbag density response function: Based on the mechanism of temperature and pressure on gas density, a nonlinear function for the gas density response of a single airbag is established. The gas density response of the i-th airbag satisfies: in, Let be the gas density of the temperature-pressure response measured in the k-th time for the i-th airbag, where T is the external temperature load and F is the external pressure load. The nonlinear mapping function for the i-th airbag is obtained by combining the gas state equation and the mechanical deformation characteristics of the airbag's flexible wall. Step 2.3: Constructing the nonlinear equation system for the array airbags: Based on the single airbag response functions of N heterogeneous airbags, integrate them to form a nonlinear equation system containing N equations, realizing the correlation between multi-dimensional density signals and temperature and pressure loads. The equation system is in the following form: ； Step 2.4: Achieve inverse solution of the mapping model: To ensure that temperature T and pressure F can be uniquely solved from the density response vector, the above nonlinear mapping relationship is optimized for effectiveness: If the application scenario involves localized, small-scale temperature and pressure changes, then the nonlinear function... Within the working interval, a first-order Taylor expansion approximation is performed to complete the local linearization process, transforming the nonlinear equation system into a directly solvable linear equation system. Y is the density response vector, and X is... Temperature-pressure vector, A is the linearization coefficient matrix; If the application scenario involves a wide range of temperature and pressure variations, multiple sets of temperature and pressure label-density response vector samples are obtained through system calibration experiments. Based on the samples, the nonlinear mapping relationship is fitted and corrected to establish a globally reversible mapping model, ensuring the uniqueness and accuracy of the inverse solution of the model throughout the entire working range.

3. A temperature-pressure measurement method based on gas density drive sensing as claimed in claim 1, wherein, The specific method for Mode 1 in step 3 is as follows: Using the nonlinear mapping model from step 2 as a framework, and training the data model with calibration data, end-to-end decoupled calculation from density signal to temperature and pressure values ​​is directly achieved. Specific steps are as follows: (1) Calibration dataset construction: The sensor airbag array is placed in a controllable temperature and pressure calibration environment, and multiple sets of discrete temperature and pressure points are selected in the entire working range. The density response vector of the airbag array was collected at each set of temperature and pressure points to construct a calibration dataset. ,in Let be the density response vector under the j-th temperature and pressure group, and m be the number of calibration samples. The number of samples must meet the requirements for suppressing overfitting during model training. (2) Data model training: Select a machine learning / deep learning model as the decoupling model and calibrate the density response vector in the dataset. For input, the corresponding temperature ,pressure To output labels, the model is trained and validated. The model structure is adjusted through hyperparameter optimization and cross-validation to ensure that the error between the model's predicted values ​​and the actual temperature and pressure values ​​meets the measurement accuracy requirements, thus obtaining the trained data-driven decoupled model. (3) Real-time decoupling calculation: Input the real-time gas density response vector collected in step 1 into the trained data-driven decoupling model. The model directly outputs the decoupled independent temperature value T and pressure value F to complete a temperature and pressure measurement.

4. The temperature-pressure measurement method based on gas density drive sensing according to claim 1, wherein, The specific method for Mode 2 in step 3 is as follows: By utilizing the characteristic differences of temperature and pressure changes in the frequency domain, the mapping model in step 2 is reconstructed in the frequency domain. Decoupling of the temperature and pressure signals is achieved through spectral separation. Specific steps are as follows: (1) Frequency domain feature calibration: In a controllable dynamic temperature and pressure environment, pure temperature dynamic excitation and pure pressure dynamic excitation are applied respectively, and the density response time domain signal under the two excitations is collected. The time domain signal is converted into a frequency domain signal by fast Fourier transform, and the temperature characteristic frequency band and pressure characteristic frequency band are calibrated. At the same time, the quantitative mapping relationship between the density signal amplitude / phase and temperature and pressure load in the two frequency bands is obtained by fitting; the low frequency band is (0.1Hz, 1Hz], and the high frequency band is (1-100Hz); (2) Real-time signal spectrum conversion: The real-time multidimensional gas density response time domain signal collected in step 1 is denoised and smoothed. Then, Fourier transform is performed on the density response signal of each airbag to convert the time domain signal into the frequency domain signal and obtain the signal amplitude and phase at each frequency point. (3) Frequency band signal separation: Based on the calibrated temperature and pressure characteristic frequency bands, a bandpass filtering algorithm is used to filter the frequency domain signal and separate the signal components within the temperature characteristic frequency band. and signal components within the pressure characteristic frequency band This eliminates the frequency domain superposition interference between the two. (4) Temperature and pressure value inversion calculation: the separated temperature frequency band signal components Pressure frequency band signal components Substitute the corresponding frequency domain quantization mapping relationship into the equations and solve for the independent temperature value T and pressure value F using the inversion algorithm. Then, fuse and average the decoupling results of the multidimensional airbag to improve the measurement accuracy.

5. The temperature-pressure measurement method based on gas density drive sensing as claimed in claim 1, wherein, The specific method for Mode 2 in step 3 is as follows: Based on the nonlinear equations from step 2, a high-dimensional eigenvector is constructed using the differentiated response characteristics of heterogeneous airbags to achieve analytical decoupled calculation of temperature and pressure loads. Specific steps include: (1) Construction of high-dimensional feature vector: Based on the gas density response vectors of N heterogeneous airbags collected in step 1, and combined with the inherent parameter differences of each airbag, feature parameters such as the ratio of density response values ​​of different airbags in steady-state response and density change rate are extracted to construct a high-dimensional decoupled feature vector. , Let be the real-time density change rate of the i-th airbag; (2) Global mapping relationship fitting: In a controllable temperature and pressure environment, high-dimensional feature vectors are collected at multiple temperature and pressure points throughout the entire working range. High-dimensional feature vectors are obtained by fitting them using a nonlinear fitting algorithm. With temperature and pressure load Global nonlinear analytical mapping relationship: This relationship is the characteristic simplified form of the nonlinear equations in step 2, which retains the core physical relationship between temperature, pressure and density. (3) Analytical decoupling calculation: The real-time gas density response vector is transformed into the corresponding high-dimensional feature vector. Substitute the obtained analytical mapping relationship into the fitting. The analytical equations are solved by numerical iterative algorithms to obtain unique temperature values ​​T and pressure values ​​F, thus completing the decoupled calculation. If the nonlinear equations have been locally linearized in step 2, the temperature and pressure values ​​can be solved analytically directly by matrix inversion, improving computational efficiency.

Images