Parameter monitoring method, device, medium, product, equipment, system and vehicle

Abstract

The invention relates to a physiological parameter monitoring method and device, a medium, a product, equipment, a system and a vehicle, and the method comprises the steps: determining physiological parameters based on a residual vibration signal collected by a residual vibration collection sensor in a vehicle seat and a physiological parameter signal collected by a physiological parameter signal collection sensor in the vehicle seat when the vehicle runs; according to the method, the residual vibration signals and the physiological parameter signals are synchronously collected through multiple non-contact sensors, the interference of mechanical vibration on physiological signal collection is effectively eliminated by combining adaptive filtering and a multi-dimensional parameter fusion technology, and the method has the advantages of being high in anti-interference capacity and high in measurement precision.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intelligent vehicle health monitoring, and in particular to a physiological parameter monitoring method, device, medium, product, equipment, system and vehicle. BACKGROUND

[0002] The existing vehicle physiological parameter monitoring technology mainly faces three technical bottlenecks:

[0003] Firstly, the optical sensing scheme has a significant environmental interference problem. The signal-to-noise ratio of the photoelectric sensor of the heart rate monitoring device based on near-infrared imaging will decrease sharply under the condition of day-night light changes or strong sunlight, especially the signal drift phenomenon caused by rapid changes in light during vehicle driving.

[0004] Secondly, the contact sensing scheme has user experience defects. The traditional electrode type sensor needs to maintain close contact with the skin, and long-term use can cause skin discomfort, and is significantly affected by the passenger's sitting posture, which seriously restricts user acceptance in long-distance driving scenarios.

[0005] Thirdly, the radio frequency sensing technology has environmental sensitivity defects. Although the millimeter wave radar scheme realizes non-contact measurement, its signal quality is easily affected by changes in temperature and humidity in the vehicle cabin, multipath effect and reflection interference of metal parts, and the measurement error can reach ±15bpm under complex road conditions.

[0006] More importantly, the existing technologies have not effectively solved the core interference source of vehicle vibration. The mechanical vibration during vehicle driving can be conducted through the seat to form a wideband noise, and its frequency spectrum range (0.1-50Hz) overlaps with the heart impact signal (0.5-5Hz). This common mode interference can cause: at the signal acquisition level, the vibration noise drowns out the effective physiological signal; at the algorithm processing level, it is difficult for traditional filtering methods to distinguish between vibration harmonics and real heart rate characteristics; at the system level, continuous mechanical vibration also affects the baseline stability of the sensor. Although some schemes try to improve by increasing hardware filtering or algorithm noise reduction, they have not fundamentally solved the interference problem of the vibration conduction path.

[0007] In view of the above problems, the existing technology needs to be improved. SUMMARY

[0008] The physiological parameter monitoring method, device, medium, product, equipment, system and vehicle provided by the embodiments of the present application have the characteristics of strong anti-interference ability, and at least partially solve the above technical problems.

[0009] In order to achieve the above purpose, according to the first aspect of the present application, a physiological parameter monitoring method is provided, which comprises:

[0010] The physiological parameter is determined based on a residual vibration signal collected by a residual vibration collection sensor in a vehicle seat and a physiological parameter signal collected by a physiological parameter signal collection sensor in the vehicle seat when the vehicle is running.

[0011] Optionally, in the foregoing method, the physiological parameter is determined based on a residual vibration signal collected by a residual vibration collection sensor in a vehicle seat and a physiological parameter signal collected by a physiological parameter signal collection sensor in the vehicle seat, and the method comprises:

[0012] An effective physiological signal is separated from the physiological parameter signal according to the residual vibration signal;

[0013] The physiological parameter is determined according to the effective physiological signal.

[0014] Optionally, in the foregoing method, the effective physiological signal is separated from the physiological parameter signal according to the residual vibration signal, and the method comprises:

[0015] An adaptive filtering method is adopted to filter out a signal associated with the residual vibration signal from the physiological parameter signal to obtain the effective physiological signal according to the residual vibration signal.

[0016] Optionally, in the foregoing method, the physiological parameter is determined according to the effective physiological signal, and the method comprises:

[0017] The effective physiological signal is subjected to band-pass filtering to obtain a filtered signal;

[0018] A time-domain candidate parameter and a frequency-domain candidate parameter are obtained based on the filtered signal;

[0019] The physiological parameter is generated according to the time-domain candidate parameter and the frequency-domain candidate parameter.

[0020] Optionally, in the foregoing method, the frequency of the band-pass filtering is 0.5 Hz to 5 Hz.

[0021] Optionally, in the foregoing method, the time-domain candidate parameter and the frequency-domain candidate parameter are obtained based on the filtered signal, and the method comprises:

[0022] A peak-to-valley detection is performed on the filtered signal to obtain a median value of a peak-to-peak interval of the filtered signal;

[0023] The time-domain candidate parameter is determined according to the median value;

[0024] The filtered signal is transformed to obtain a power spectrum of the filtered signal;

[0025] The frequency-domain candidate parameter is determined according to a spectrum peak of the power spectrum.

[0026] Optionally, the generating the physiological parameter according to the time domain candidate parameter and the frequency domain candidate parameter in the foregoing method comprises:

[0027] performing Kalman fusion on the time domain candidate parameter and the frequency domain candidate parameter to obtain the physiological parameter.

[0028] Optionally, the foregoing method further comprises:

[0029] controlling the vehicle seat to actively reduce vibration based on a seat vibration signal collected by a vibration sensor in the vehicle seat and the residual vibration signal when the vehicle is running.

[0030] Optionally, the controlling the vehicle seat to actively reduce vibration based on a seat vibration signal collected by a vibration sensor in the vehicle seat and the residual vibration signal in the foregoing method comprises:

[0031] generating a reverse vibration reduction signal based on the seat vibration signal;

[0032] controlling an actuator in the vehicle seat to work according to the reverse vibration reduction signal;

[0033] adjusting the work of the actuator based on the residual vibration signal.

[0034] Optionally, the adjusting the work of the actuator based on the residual vibration signal in the foregoing method comprises:

[0035] adjusting a filter coefficient based on the residual vibration signal and a convergence coefficient;

[0036] adjusting the reverse vibration reduction signal according to the filter coefficient and the seat vibration signal.

[0037] Optionally, the foregoing method further comprises:

[0038] determining a physiological parameter when the vehicle is running and a user is present in the vehicle seat.

[0039] Optionally, the foregoing method further comprises:

[0040] determining the physiological parameter signal according to a signal quality of each physiological parameter signal collection sensor and a collection signal of each physiological parameter signal collection sensor.

[0041] Optionally, the determining the physiological parameter signal according to a signal quality of each physiological parameter signal collection sensor and a collection signal of each physiological parameter signal collection sensor in the foregoing method comprises:

[0042] determining a weight of each physiological parameter signal collection sensor according to a signal quality of each physiological parameter signal collection sensor;

[0043] The physiological parameter signals are obtained by processing the acquired signals from each physiological parameter signal acquisition sensor according to the weight of each sensor.

[0044] Optionally, the aforementioned method further includes:

[0045] When a change in attitude is detected, a smooth transition method is used to switch the signal source.

[0046] According to a second aspect of this application, a physiological parameter monitoring device is provided, comprising:

[0047] The monitoring module is used to determine physiological parameters based on residual vibration signals collected by residual vibration acquisition sensors in the vehicle seat and physiological parameter signals collected by physiological parameter signal acquisition sensors in the vehicle seat when the vehicle is in motion.

[0048] According to a third aspect of this application, a computer-readable storage medium is provided, on which a computer program is stored, which, when executed by a processor, implements the above-described physiological parameter monitoring method.

[0049] According to a fourth aspect of this application, a computer program product is provided, comprising a computer program that, when executed by a processor, implements the above-described physiological parameter monitoring method.

[0050] According to a fifth aspect of this application, an electronic device is provided, comprising: a memory having a computer program stored thereon; and a processor for executing the computer program in the memory to implement the above-described physiological parameter monitoring method.

[0051] According to a sixth aspect of this application, a physiological parameter monitoring system is provided, comprising:

[0052] A signal-to-noise ratio (SNR) enhancement device, installed inside a vehicle seat, includes a signal isolation pad, a residual vibration acquisition sensor located on one side of the signal isolation pad, and a physiological parameter signal acquisition sensor located on the other side of the signal isolation pad; and

[0053] Physiological parameter measuring device;

[0054] in:

[0055] The residual vibration acquisition sensor is configured to acquire residual vibration signals when the vehicle is in motion.

[0056] The physiological parameter signal acquisition sensor is configured to acquire physiological parameter signals when the vehicle is in motion.

[0057] The physiological parameter measuring device is configured to determine physiological parameters based on the residual vibration signal and the physiological parameter signal.

[0058] Optionally, the aforementioned system further includes:

[0059] An active damping device, installed within the vehicle seat, includes a vibration sensor, a vibration source, and an actuator; wherein:

[0060] The vibration sensor is configured to collect seat vibration signals when the vehicle is in motion.

[0061] The vibration source is configured to control the actuator to perform active vibration reduction based on the seat vibration signal and the residual vibration signal.

[0062] Optionally, the aforementioned system further includes:

[0063] The monitoring module is configured to enable the physiological parameter monitoring system when the vehicle is in motion and a user is present in the vehicle seat.

[0064] According to the seventh aspect of this application, a vehicle is provided, including the aforementioned electronic device or the aforementioned physiological parameter monitoring system.

[0065] This application embodiment uses multiple non-contact sensors to simultaneously collect residual vibration signals and physiological parameter signals while the vehicle is in motion. Combined with adaptive filtering and multi-dimensional parameter fusion technology, it effectively eliminates the interference of mechanical vibration on the collection of physiological signals, and has the advantages of strong anti-interference ability, high measurement accuracy, and good user experience.

[0066] Other features and advantages of this application will be described in detail in the following detailed description section. Attached Figure Description

[0067] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0068] To gain a more complete understanding of this application and its beneficial effects, the following description will be provided in conjunction with the accompanying drawings, wherein the same reference numerals in the following description denote the same parts.

[0069] Figure 1 This is a flowchart of a physiological parameter monitoring method provided in an embodiment of this application;

[0070] Figure 2 This is a schematic diagram of a system architecture for a physiological parameter monitoring system provided in an embodiment of this application;

[0071] Figure 3This is a schematic diagram of a seat provided in an embodiment of this application;

[0072] Figure 4 This is a schematic diagram illustrating one possible arrangement of sensors in a seat according to an embodiment of this application;

[0073] Figure 5 This is a schematic diagram illustrating the effect of the active damping algorithm provided in an embodiment of this application;

[0074] Figure 6 This is a schematic diagram of an active damping algorithm provided in an embodiment of this application;

[0075] Figure 7 This is a schematic diagram of an adaptive filtering method provided in an embodiment of this application;

[0076] Figure 8 This is a schematic diagram illustrating the effect of the adaptive filtering method provided in an embodiment of this application;

[0077] Figure 9 This is a schematic diagram of a filtering signal processing method provided in an embodiment of this application;

[0078] Figure 10 This is a schematic diagram of a vehicle provided in an embodiment of this application. Detailed Implementation

[0079] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the protection scope of this application.

[0080] In traditional vehicle physiological parameter monitoring systems, non-contact sensors are susceptible to signal quality degradation due to driving vibrations. Mechanical vibrations are transmitted through the seat to the physiological signal acquisition sensors, causing aliasing with the target physiological signal in the frequency domain, resulting in complex noise pollution. Time-varying vibration noise makes it difficult for bandpass filters in the signal preprocessing stage to effectively separate effective frequency bands. Time-domain feature extraction algorithms are susceptible to the effects of transient vibration shocks, leading to spurious peak detection. Frequency-domain power spectrum analysis suffers from decreased peak identification accuracy due to the increased noise floor. Such interference directly affects the time-frequency domain parameter fusion effect of the heart rate calculation module, causing periodic fluctuations and steady-state errors in the physiological parameter output values.

[0081] For example, in a cockpit monitoring system equipped with millimeter-wave radar, when a vehicle travels at 80 km / h on a paved road, the second-order vibration component of the engine and the tire-excited vibration form a composite vibration spectrum at the seat, with its energy concentrated in the 1.5-3 Hz frequency band. This frequency band significantly overlaps with the typical adult resting heart rate (60-100 bpm corresponds to 1-1.67 Hz), resulting in vibration noise accounting for more than 40% of the phase change sequence after demodulation of the millimeter-wave signal. When the signal processing unit performs a fast Fourier transform, the vibration noise forms multiple spurious spectral peaks in the target frequency band, causing an instantaneous error of ±8 bpm in the peak detection algorithm. Simultaneously, the sensor displacement caused by seat vibration leads to multipath effects in the millimeter-wave echo signal, further reducing the signal-to-noise ratio of the original signal to below 3 dB.

[0082] If the above issues are not addressed, the physiological parameter monitoring system will fail to meet medical-grade accuracy requirements, leading to the failure of health warning functions. In continuous monitoring scenarios, the accumulation of vibration noise may cause nonlinear distortion in the signal processing link, resulting in abnormal fluctuations in the heart rate trend curve. When vehicle driving conditions change, the dynamic migration of the vibration spectrum will exacerbate the convergence difficulty of the adaptive filtering algorithm, extending the system response time to over a second. These technical deficiencies will severely restrict the reliability of in-vehicle health monitoring systems in autonomous driving scenarios, failing to meet the technical requirements of the ISO 13482 standard for mobile physiological monitoring devices.

[0083] Faced with the aforementioned challenges, this application first considers how to establish a vibration and noise reference standard while maintaining the advantages of non-contact monitoring. Traditional methods attempt to suppress vibration interference through hardware isolation or sensor array arrangements, but these methods suffer from increased cost and structural complexity. Further analysis reveals that seat vibration exhibits spatiotemporal correlation characteristics along its transmission path, and a modelable transmission relationship exists between the residual vibration signal and the noise component in the physiological signal. This leads to the technical concept of constructing dual signal acquisition channels to capture the original physiological signal and the residual vibration reference signal respectively, thereby establishing a dynamic noise cancellation model.

[0084] In response, this application proposes a method for monitoring physiological parameters, such as... Figure 1 As shown, in some embodiments, the physiological parameter monitoring method provided in this application includes the following steps:

[0085] S100: When the vehicle is in motion, physiological parameters are determined based on the residual vibration signals collected by the residual vibration acquisition sensor in the vehicle seat and the physiological parameter signals collected by the physiological parameter signal acquisition sensor in the vehicle seat.

[0086] Among them, the residual vibration acquisition sensor and the physiological parameter signal acquisition sensor acquire data in a non-contact manner. The non-contact manner refers to the acquisition of physiological parameter signals through sensing technology that does not require physical contact. Physiological parameters include heart rate, etc. Specifically, cardiac impact signal sensors or piezoelectric film sensors can be used. This method avoids the impact of contact sensors on the user's physical sensation and avoids the defect of optical sensors being easily interfered with by ambient light.

[0087] Among them, the residual vibration acquisition sensor refers to the device installed in the vehicle seat structure to detect the residual vibration. Specifically, it can be implemented by using an accelerometer or a piezoelectric vibration sensor. Its function is to capture the residual vibration signal after vibration reduction treatment and provide a noise reference for subsequent signal separation.

[0088] Among them, the physiological parameter signal acquisition sensor refers to the device integrated on the surface of the seat to acquire human physiological characteristic signals. Specifically, it can be implemented by using a cardiac impactor sensor or a flexible piezoelectric sensor array. Its function is to acquire a mixed signal containing real physiological information and mechanical vibration noise.

[0089] Among them, "when the vehicle is in motion" refers to the operating conditions under which the vehicle is in motion as determined by the vehicle speed sensor or vehicle status monitoring module. Specifically, this can be achieved by parsing CAN bus signals or using inertial measurement unit data. This time constraint ensures that the technical solution activates the monitoring function for the operating environment where vibration interference is most significant.

[0090] The core innovation of this application lies in the establishment of a dual-source signal collaborative processing mechanism. Through the combined action of non-contact physiological signal acquisition and residual vibration monitoring, physiological parameters can be effectively extracted while the vehicle is in motion. This mechanism utilizes the residual vibration signal as a noise reference standard, combines the time-frequency characteristics of the physiological mixed signal, and eliminates mechanical vibration interference through signal separation technology, thus solving the technical problem of insufficient signal-to-noise ratio of a single sensor in complex vibration environments.

[0091] The working process and principle of this application are as follows: During vehicle operation, residual vibration signals and physiological parameter signals from the seat are collected non-contactly. The residual vibration sensor is located at the bottom of the seat to capture residual vibrations after seat damping; the physiological parameter signal sensor is integrated into the seat back, acquiring physiological signals including heart rate without direct contact with the occupant. These two signals are amplified and preliminarily filtered by signal conditioning circuits before being input to the signal processing unit. The signal processing unit first performs spectral analysis on the residual vibration signal to identify the main vibration frequency components. Then, using an adaptive filtering algorithm, with the residual vibration signal as a reference input, it separates vibration-related noise components from the physiological parameter signal. The resulting purified physiological signal is then processed by a bandpass filter to extract the heart rate-related frequency band.

[0092] Next, the signal processing unit performs peak detection in the time domain, calculates the time interval between adjacent peaks, and converts it into an instantaneous heart rate value. Simultaneously, it performs power spectrum analysis in the frequency domain to find the heart rate corresponding to the dominant frequency component. Finally, a Kalman filter is used to fuse the heart rate estimation results from the time and frequency domains to obtain the final heart rate output.

[0093] Throughout the process, the introduction of residual vibration signals provides a dynamic reference for noise cancellation, enabling the system to adapt to changes in vibration characteristics under different road conditions. The non-contact acquisition method ensures a good user experience, allowing for continuous monitoring of physiological parameters during daily driving. Through multi-level signal processing and fusion strategies, the accuracy of heart rate detection in vibration environments is effectively improved.

[0094] As a preferred embodiment, the solution of this application is specifically implemented as follows:

[0095] In a moving car, piezoelectric accelerometers are installed under the seats to collect residual vibrations, while capacitive sensors are built into the seat backs to collect physiological parameter signals. The system automatically activates when it detects that an occupant has taken their seat and the vehicle has started moving.

[0096] The residual vibration sensor continuously collects seat vibration acceleration data at a sampling rate of 200Hz. The physiological parameter signal sensor collects minute deformation signals of the area where the occupant's back contacts the seat at a sampling rate of 1000Hz. Both signals are then processed by a microcontroller based on an ARM Cortex-M4 core after passing through their respective preamplifiers and 16-bit analog-to-digital converters.

[0097] The microcontroller first performs a Fast Fourier Transform on the residual vibration signal from the most recent 10 seconds to identify the main vibration frequency components. Then, using a Least Mean Square Error (LMS) adaptive filtering algorithm, the residual vibration signal is used as a reference input to separate vibration noise from the physiological parameter signal. The filter coefficients are updated every 50ms to adapt to changes in vibration characteristics.

[0098] The purified physiological signal is passed through a bandpass filter of 0.7-2.5 Hz to extract the heart rate-related frequency band. Then, a peak detection algorithm is used in the time domain to calculate the time interval between adjacent peaks and convert it into an instantaneous heart rate value. Simultaneously, power spectral density estimation is performed on the signal from the most recent 30 seconds to identify the dominant frequency component within the 0.7-2.5 Hz range, which is used as the frequency domain heart rate estimate.

[0099] Finally, a Kalman filter is used to fuse the heart rate estimates in the time and frequency domains. The filter's state transition model assumes that the heart rate changes slowly over a short period, while the observation model considers the different characteristics of the time and frequency domain estimates. The fused heart rate is updated once per second and output to the in-vehicle display screen.

[0100] Through the above-described solution, this application effectively addresses the accuracy issue of physiological parameter monitoring while the vehicle is in motion. By introducing residual vibration signals as a noise reference, the system can dynamically adapt to vibration characteristics under different road conditions, significantly improving vibration and noise suppression. The non-contact sensor arrangement avoids the impact of traditional contact sensors on user comfort, enabling continuous physiological parameter monitoring during daily driving. The application of multi-level signal processing and fusion strategies further enhances the stability and reliability of heart rate detection. Therefore, this solution achieves high-precision physiological parameter monitoring while the vehicle is in motion, ensuring a positive user experience and providing reliable data support for in-vehicle health management systems.

[0101] In some of the solutions described above in this application, physiological parameters and residual vibration signals are collected based on non-contact sensors to determine physiological parameters. However, during vehicle operation, the physiological parameter signals are mixed with seat vibration noise signals. As a result, when physiological parameters are calculated directly using the original physiological parameter signals, the vibration interference component will reduce the signal-to-noise ratio of the effective physiological signal, thereby affecting the accuracy of physiological parameter detection.

[0102] In response, this application further proposes a signal separation method, including: determining physiological parameters based on residual vibration signals collected by residual vibration acquisition sensors in the vehicle seat and physiological parameter signals collected by physiological parameter signal acquisition sensors, including separating effective physiological signals from physiological parameter signals based on residual vibration signals, and determining physiological parameters based on effective physiological signals.

[0103] The residual vibration signal is used as a reference source for vibration noise, and signal separation is achieved through adaptive filtering. The residual vibration signal and the vibration interference component in the physiological parameter signal have a time-domain correlation. By constructing an adaptive filter model, the residual vibration signal is used as a reference input, and the filter coefficients are adjusted in real time to cancel the vibration noise in the physiological parameter signal. The filter coefficient update uses the least mean square algorithm, and the convergence coefficient can be set to a value in the range of 0.01 to 0.1, for example, 0.05, to balance convergence speed and stability. When the effective physiological signal is processed by bandpass filtering, the passband range is set to 0.5Hz to 5Hz, covering the frequency corresponding to typical heart rate. Time-domain candidate parameters are calculated by detecting the median of the interpeak sequence. The buffer capacity N can be set to 10 to 20 cycles of data, and smoothing filtering uses a threshold of 1.2 times the mean and 0.8 times the threshold to remove outliers. Frequency-domain candidate parameters obtain the dominant frequency component through power spectrum analysis, and the FFT transform window length can be set to 5 to 10 seconds. The time-domain and frequency-domain results are fused using Kalman filtering. The process noise covariance matrix parameter can be configured from 0.1 to 1.0, and the observation noise covariance matrix parameter is set from 0.5 to 2.0.

[0104] Specifically, the physiological parameter signal acquisition sensor and the residual vibration acquisition sensor are synchronously triggered, with a sampling rate set to 100Hz to 1000Hz. The residual vibration signal is input into an adaptive filter to generate a predictive noise component. This component is subtracted from the mixed signal to obtain the effective physiological signal. The bandpass-filtered signal is analyzed in the time domain using a sliding window mechanism, with a window length configurable from 3 to 5 seconds and a step size of 1 second. Peak detection employs the local extremum method, with the interval between adjacent peaks limited to 0.3 to 2 seconds to eliminate interference. Frequency domain analysis uses a Hanning window function to reduce spectral leakage, with the peak search range limited to 0.8Hz to 3.5Hz, corresponding to heart rates of 48 bpm to 210 bpm. The state equation of the Kalman filter is established as a heart rate change model. By fusing the time-domain median HR1 and the frequency-domain peak HR2, the final heart rate value HR_f is output. This dual-modal verification mechanism ensures the continuity of heart rate output even when a single detection method is affected by sudden interference. The effective physiological signal-to-noise ratio of the separated signals can be improved by 6dB to 10dB, and the heart rate calculation error is controlled within ±2bpm.

[0105] As a preferred embodiment, the solution of this application is specifically implemented as follows:

[0106] Physiological parameters are determined based on residual vibration signals collected by residual vibration sensors in the vehicle seat and physiological parameter signals collected by physiological parameter signal sensors in the vehicle seat. Specifically, firstly, effective physiological signals are separated from the physiological parameter signals based on the residual vibration signals. Further, physiological parameters are determined based on the effective physiological signals.

[0107] Through the above technical solution, this application can effectively separate the effective physiological signal from the mixed signal, reducing the impact of vehicle vibration on the detection of physiological parameters. By using the residual vibration signal as a reference, the changing characteristics of vibration noise are dynamically tracked, and its contamination of the physiological signal is specifically eliminated, thereby improving the accuracy of physiological parameter detection. In addition, the step-by-step processing method avoids the direct impact of residual noise on the final parameter calculation, further enhancing the reliability of physiological parameter detection.

[0108] In some of the solutions described above in this application, traditional adaptive filtering methods have difficulty dynamically adjusting filter parameters, resulting in incomplete separation of vibration-related signals. Periodic vibration interference components still remain in the effective physiological signals, affecting the accuracy of subsequent physiological parameter calculations.

[0109] In response, this application further proposes a separation scheme: using an adaptive filtering method, based on the residual vibration signal, the associated signal of the residual vibration signal is separated from the physiological parameter signal to obtain the effective physiological signal.

[0110] The effective cardiac impact signal acquired by the cardiac impact signal acquisition sensor is denoted as s1, and the acquired seat vibration signal is denoted as n1; the residual seat vibration signal acquired by the seat vibration sensor is denoted as n2. The cardiac impact signal e(n) obtained by the adaptive filter is realized through the error calculation formula e(n)=s1+n1-y, where y is the output of n2 after passing through the filter W, and its calculation formula is y=n2*w. The filter coefficient update is realized through W=Wp-c*e*n2, where c is the convergence coefficient of the adaptive filter, and Wp is the filter coefficient of the previous time step.

[0111] The error signal e(n) is constructed by using the difference between the seat vibration interference n1 and the filter output y as the basis for dynamic adjustment. The filter output y is generated by convolving the residual vibration signal n2 with the filter coefficients w, which can be replaced by a weighted summation of the finite impulse response filter. During the filter coefficient update process, the convergence coefficient c can be set to a fixed value in the range of 0.001 to 0.1, or dynamically adjusted according to the signal energy. The inheritance mechanism of the filter coefficients Wp from the previous moment can be implemented through register storage or a circular buffer to ensure the temporal continuity of parameter updates.

[0112] Specifically, when the residual vibration signal n2 is input to the adaptive filter, it is linearly combined with the current filter coefficients w to generate the output signal y. This output signal is compared with the vibration interference component n1 in the original signal in the error calculation module to generate the error signal e(n). The product of this error signal and the reference signal n2 is scaled by the convergence coefficient c to form the incremental adjustment of the filter coefficients. By updating the filter coefficients in real time, the output signal y gradually approximates the real vibration interference n1, thus retaining the effective core impact signal s1 in the error signal e(n). In this process, the value of the convergence coefficient c directly affects the parameter adjustment step size. A larger c value can speed up the convergence speed but may cause parameter oscillations, while a smaller c value improves stability but prolongs the convergence time. By setting the c value to an empirical value in the range of 0.01 to 0.05, a balance can be achieved between convergence speed and stability. The historical inheritance mechanism of the filter coefficients ensures that each parameter update is based on the previous optimization, avoiding interference from parameter mutations in the signal separation process. This dynamic adjustment mechanism enables the filter to track the spectral characteristics of seat vibration in real time, effectively solving the problem of performance degradation of traditional fixed-parameter filters in complex vehicle vibration environments.

[0113] As a preferred embodiment, the solution of this application is specifically implemented as follows:

[0114] An adaptive filtering method is employed to separate the associated signal of the residual vibration signal from the physiological parameter signal, thereby obtaining the effective physiological signal. Specifically, the cardiac impact signal acquisition sensor acquires the effective cardiac impact signal s1 and the seat vibration signal n1, while the residual seat vibration sensor acquires the seat vibration signal n2. The cardiac impact signal e(n) obtained by the adaptive filter is calculated as e(n) = s1 + n1 - y, where y is the output of n2 after passing through filter W, calculated as y = n2 * w. The update formula for filter W is W = Wp - c * e * n2, where c is the convergence coefficient of the adaptive filter, and Wp is the filter coefficient at the previous time step.

[0115] Furthermore, the sampling frequency of the cardiac impact signal acquisition sensor and the residual seat vibration sensor can be set to 100Hz. The filter W can be initialized as a zero vector of length 32. The convergence coefficient c can be set to 0.01. At each sampling time, the filter output y is calculated first, then the error signal e(n) is calculated, and finally the filter coefficients W are updated. This process iterates until the energy of the error signal e(n) decreases below a preset threshold, for example, 1% of the original signal energy.

[0116] Therefore, by dynamically adjusting the filter parameters, the complexity and time-varying nature of seat vibration during vehicle operation can be adapted to achieve accurate separation of vibration-related signals.

[0117] Through the above technical solution, this application can dynamically adjust the filter parameters to effectively adapt to the complexity and time-varying nature of seat vibration during vehicle operation. This adaptive filtering method can more thoroughly separate vibration-related signals and significantly reduce the residual periodic vibration interference components in the effective physiological signals. This improves the accuracy of subsequent physiological parameter calculations and provides a more reliable signal basis for monitoring physiological parameters while the vehicle is in motion.

[0118] In some of the solutions described above in this application, since there may still be residual noise interference or inconsistencies in time-frequency domain features in the effective physiological signal, direct feature extraction based on a single domain may lead to unstable or inaccurate physiological parameter calculation results. It is necessary to combine the complementary advantages of time-domain and frequency-domain features to improve the robustness of parameter estimation.

[0119] In response, this application further proposes specific separation schemes, including:

[0120] The effective physiological signal is bandpass filtered to obtain the filtered signal; time-domain candidate parameters and frequency-domain candidate parameters are obtained based on the filtered signal; physiological parameters are generated based on the time-domain candidate parameters and frequency-domain candidate parameters.

[0121] The bandpass filter frequency range is set to 0.5Hz to 5Hz, which can cover the typical frequency band of human heart rate, thereby suppressing high-frequency noise and low-frequency motion artifacts.

[0122] The process of obtaining time-domain candidate parameters and frequency-domain candidate parameters based on the filtered signal includes: performing peak and trough detection on the filtered signal to obtain the median of the peak-to-peak interval of the filtered signal; determining the time-domain candidate parameters based on the median; transforming the filtered signal to obtain the power spectrum of the filtered signal; and determining the frequency-domain candidate parameters based on the spectral peaks of the power spectrum.

[0123] The process of generating the physiological parameters based on the time-domain candidate parameters and the frequency-domain candidate parameters includes: performing Kalman fusion on the time-domain candidate parameters and the frequency-domain candidate parameters to obtain the physiological parameters.

[0124] Specifically, in the generation of time-domain candidate parameters, peak and trough detection is applied to the filtered signal. This is achieved by searching for maxima within a certain time window and recording their occurrence times. When the number of maxima in the buffer reaches or exceeds N, the peak-to-peak interval sequence is calculated. Abnormal intervals are eliminated through mean range filtering; for example, intervals less than 0.8 times or more than 1.2 times the mean are filtered out. The median of the remaining interval sequence is selected and converted into a time-domain heart rate value. In the generation of frequency-domain candidate parameters, the filtered signal is subjected to FFT transformation to obtain the power spectrum. The peak search range is limited to 0.5Hz to 5Hz, and the dominant frequency component is mapped to a frequency-domain heart rate value. The time-domain and frequency-domain candidate parameters are dynamically fused using the Kalman filtering algorithm, where the anti-frequency offset capability of the time-domain parameters complements the noise resistance capability of the frequency-domain parameters.

[0125] Specifically, the effective physiological signal is first bandpass filtered, for example, using a cutoff frequency of 0.5Hz to 5Hz to preserve the signal components in the heart rate-related frequency band. The filtered signal is then processed in two paths: the time domain and the frequency domain. In the time domain path, peak and trough detection is dynamically performed, and the occurrence of maxima is saved to a buffer. When the amount of data in the buffer reaches a preset threshold N, the peak-to-peak interval sequence is calculated, and outliers are filtered out using the mean range. The median of the latest M effective intervals is converted into the time-domain heart rate value HR1. In the frequency domain path, signal segments exceeding the time threshold tn are truncated and subjected to FFT transformation. The main peak frequency of the power spectrum is extracted and converted into the frequency-domain heart rate value HR2. HR1 and HR2 are input into a Kalman filter, where the stability of the time-domain parameters and the global consistency of the frequency-domain parameters are fused through dynamic weight adjustment, ultimately generating a more robust physiological parameter HR_f. Through the synergistic effect of time-domain anomaly interval elimination and frequency-domain spectral peak search, the influence of transient interference and local signal distortion is effectively suppressed. At the same time, the adaptive fusion mechanism of Kalman filtering further improves the parameter estimation accuracy under complex interference scenarios.

[0126] As a preferred embodiment, the solution of this application is specifically implemented as follows:

[0127] The effective physiological signal is bandpass filtered to obtain the filtered signal. The frequency range of the bandpass filter is 0.5Hz to 5Hz. Further, time-domain and frequency-domain candidate parameters are obtained based on the filtered signal. Specifically, peak and trough detection is performed on the filtered signal to search for the maximum value within a certain time period, and the time tx when the maximum value occurs is saved. tx is saved to a buffer. When the number of results saved in the buffer is greater than or equal to N, the difference between the buffer results is calculated to obtain the peak-to-peak interval (PS) of the cardiac impulse signal. The PS sequence is smoothed by filtering, removing intervals less than 0.8 times the mean and more than 1.2 times the mean, resulting in the filtered peak-to-peak interval sequence PSF. The median RR_med of the latest M peak-to-peak intervals in the PSF sequence is calculated, and the time-domain candidate parameter HR1 is obtained using the heart rate calculation formula HR = 60 / RR_med.

[0128] An FFT transformation is performed on the cardiac impact signal acquired at a time exceeding tn to obtain its power spectrum. A spectral peak in the power spectrum ranging from 0.5 Hz to 5 Hz is searched, and the obtained peak hf is substituted into the heart rate calculation formula Hr = 60 * hf to obtain the frequency domain candidate parameter HR2.

[0129] Finally, Kalman heart rate fusion was performed on the time-domain candidate parameter HR1 and the frequency-domain candidate parameter HR2 to obtain the physiological parameter HR_f.

[0130] Through the above technical solutions, this application can effectively improve the robustness of physiological parameter estimation. Therefore, by combining the complementary advantages of time-domain and frequency-domain features, it solves the problem that direct feature extraction based on a single domain may lead to unstable or inaccurate physiological parameter calculation results. Specifically, bandpass filtering selectively preserves heart rate-related frequency band signals and suppresses high-frequency noise and low-frequency motion artifacts. Time-domain processing eliminates the fluctuations in heart rate calculation caused by transient interference through peak and trough detection, abnormal interval removal, and median calculation. Frequency-domain processing extracts the dominant frequency component through FFT power spectrum analysis, avoiding errors caused by local distortions in the time-domain signal. Finally, Kalman filtering dynamically weights and fuses candidate parameters in the time and frequency domains, combining the advantages of both to generate more robust heart rate parameters. Furthermore, time-domain processing achieves dynamic smoothing through a buffer mechanism and median filtering, while frequency-domain processing ensures global consistency through spectral peak search within a fixed time window. The complementary nature of these two processes, combined with the adaptive adjustment of weights using the Kalman algorithm, forms an adaptive optimization for complex interference scenarios.

[0131] In some of the solutions mentioned above in this application, although the interference problem of contact measurement is solved by collecting physiological parameter signals based on non-contact sensors, the seat vibration when the vehicle is in motion will still reduce the accuracy of physiological parameter detection through residual vibration signals. At the same time, continuous vibration affects riding comfort. Existing solutions cannot reduce the dual negative impact of vibration energy from the source by only processing signals.

[0132] In response, this application further proposes a technical solution for controlling the seat to perform active vibration reduction based on seat vibration signals and residual vibration signals when the vehicle is in motion.

[0133] The active vibration damping mechanism incorporates the following core features: real-time acquisition of seat vibration signals and generation of reverse signals can be achieved through a phase inversion algorithm, such as generating a corresponding reverse waveform based on the frequency and amplitude of the vibration signal; the actuator control can employ linear motors or piezoelectric ceramic actuators, with response times controlled in milliseconds to match the vibration frequency; closed-loop adjustment of residual vibration signals can be achieved through adaptive filter coefficient updates, for example, setting the convergence coefficient range to 0.001 to 0.1 to balance adjustment speed and stability. The coordinated processing of vibration sources and signal interference is reflected in the fact that the reverse damping signal directly reduces the seat vibration energy, while the residual signal feedback optimizes actuator parameters; both work together to reduce the interference intensity of vibration on physiological signal acquisition sensors.

[0134] Specifically, when the vehicle is in motion, the vibration sensor collects the seat vibration signal, extracts the dominant vibration frequency component through real-time spectrum analysis, and generates a reverse signal with a 180-degree phase difference to drive the actuator to produce a canceling vibration. After the actuator output signal is superimposed on the seat vibration signal, the residual vibration energy is reduced. The residual vibration sensor continuously monitors the residual vibration after damping. When it detects that the energy in a specific frequency band exceeds a threshold, for example, energy exceeding 0.1 m / s² in the 5Hz to 20Hz range... 2 This triggers the filter coefficient update mechanism. The filter weights are adjusted using a least mean square algorithm, allowing the actuator's output reverse signal to dynamically match the current vibration characteristics. During this process, the periodic characteristics of the micro-periodic vibration signal are used to correct signal transmission delays; for example, for vibrations with a period of 50ms, a 10ms phase compensation is set to improve cancellation efficiency. This closed-loop control mechanism enables the damping system to adapt to vibration changes under different road conditions, effectively controlling the seat vibration acceleration within the human comfort threshold of 0.5m / s² while reducing interference from physiological signal acquisition. 2 the following.

[0135] As a preferred embodiment, the solution of this application is specifically implemented as follows:

[0136] During vehicle operation, vibration sensors in the vehicle seat collect seat vibration signals in real time. Simultaneously, residual vibration sensors collect residual vibration signals. The control unit receives these two signals and processes and analyzes them. Based on the analysis results, the control unit generates a reverse damping signal, whose phase is opposite to the original vibration signal. The control unit sends the reverse damping signal to the actuator in the vehicle seat. The actuator generates a counteracting vibration based on the received reverse damping signal, thereby achieving active damping.

[0137] Furthermore, the control unit continuously monitors changes in the residual vibration signal. If the detected residual vibration signal exceeds a preset threshold, the control unit dynamically adjusts the parameters of the reverse damping signal, such as amplitude and frequency, to optimize the damping effect. This closed-loop control mechanism ensures that the active damping system can adapt to vibration characteristics under different road conditions and vehicle speeds.

[0138] Therefore, the technical solution of this application can not only effectively reduce the interference of vehicle vibration on physiological parameter monitoring, but also significantly improve the comfort of passengers. Specifically, through active damping, the vibration amplitude of the seat can be greatly reduced, thereby reducing the impact of vibration on non-contact physiological parameter sensors and improving the accuracy of physiological parameter monitoring. At the same time, the reduced vibration intensity also alleviates the fatigue caused by prolonged vibration for passengers, improving the overall riding experience.

[0139] Through the above technical solution, this application achieves a collaborative solution to vehicle vibration problems from both the vibration source and signal interference levels. The active damping system acquires vibration characteristics in real time and generates a counteracting signal with opposite phase, which directly acts on the seat actuator to cancel the original vibration energy, thereby reducing the interference intensity of vibration on the physiological signal acquisition sensors. Simultaneously, by monitoring the residual vibration signal that is not completely canceled after damping, the generation parameters of the reverse damping signal are dynamically optimized, forming a closed-loop control mechanism. This dual signal feedback mechanism not only improves the real-time performance and adaptability of active damping but also reduces vibration interference with physiological parameter signals while minimizing fatigue caused by continuous vibration, achieving a dual improvement in the accuracy of physiological parameter monitoring and ride comfort.

[0140] In some of the solutions mentioned above in this application, an active vibration reduction method is proposed to generate a reverse damping signal based on the seat vibration signal to control the operation of the actuator. However, due to the dynamic change characteristics of the vibration signal when the vehicle is in motion, it is difficult to match the phase and amplitude changes of the vibration in real time by relying solely on a fixed reverse damping signal generation method. As a result, the residual vibration signal cannot be effectively suppressed, which in turn affects the stability and continuity of the damping effect.

[0141] In response, this application further proposes a specific optimization scheme: generating a reverse damping signal based on the seat vibration signal, since the vehicle vibration signal is a periodic signal, the delay of the reverse damping signal can be corrected based on the periodicity; controlling the actuator in the vehicle seat to work according to the reverse damping signal; and adjusting the actuator operation based on the residual vibration signal.

[0142] The periodic delay correction can be achieved through a phase compensation algorithm, such as using a dynamic phase tracking module to calculate the period length of the vibration signal in real time and advance or delay the generation time of the reverse signal by an integer multiple of the corresponding period. The actuator control circuit can be configured with a multi-level drive mode, for example, using full power output to quickly suppress vibration in the initial stage and switching to pulse width modulation mode to reduce power consumption in the stable stage. The adjustment mechanism for the residual vibration signal can include an adaptive filtering algorithm, such as using the energy value of the residual signal as a feedback parameter and iteratively updating the filter coefficients using a least mean square algorithm.

[0143] Specifically, after the seat vibration signal is collected by the vibration sensor, the dominant vibration frequency and its phase information are first extracted by the period detection module. For example, a fast Fourier transform is used to identify the dominant frequency component in the range of 0.5-20Hz. During the generation of the reverse damping signal, the signal delay is dynamically corrected by a phase compensator. For example, when a 10Hz vibration component is detected, the phase of the reverse signal is advanced to correspond to 1 / 4 of the phase angle of a 0.1-second period. After receiving the corrected reverse signal, the actuator generates reverse mechanical vibration through an electromagnetic actuator. The operating frequency range of the actuator can be set to 5-100Hz to cover common vehicle vibration frequency bands. The residual vibration sensor continuously monitors the damping effect. When the residual vibration energy exceeds a threshold, for example, a threshold set at 15% of the initial vibration energy, a filter coefficient update mechanism is triggered. The amplitude ratio and phase offset of the reverse signal are adjusted through a gradient descent algorithm, ultimately causing the residual vibration energy to converge to the target range.

[0144] As a preferred embodiment, the solution of this application is specifically implemented as follows:

[0145] A reverse damping signal is generated based on the seat vibration signal. Since vehicle vibration signals are periodic, the delay of the reverse damping signal can be corrected based on this periodicity. Specifically, an adaptive filtering algorithm can be used to analyze the seat vibration signal, extract its periodic characteristics, and dynamically adjust the phase and amplitude of the reverse damping signal based on these characteristics. For example, the minimum mean square error (LMS) algorithm can be used to update the filter coefficients in real time to generate a reverse damping signal that is out of phase and matches the amplitude of the seat vibration signal.

[0146] The actuators in the vehicle seat are controlled based on the reverse damping signal. The generated reverse damping signal can be converted into a drive voltage or current, which acts on the electromagnetic or hydraulic actuators in the seat. The actuators generate a movement opposite to the seat vibration according to the drive signal, thereby counteracting the seat vibration.

[0147] Based on the residual vibration signal, the actuator operation is adjusted. This can be achieved by installing an accelerometer on the seat to monitor the residual vibration signal in real time. The residual vibration signal is then fed back into the control algorithm to further optimize the generation of the reverse damping signal. For example, an adaptive control strategy can be employed to dynamically adjust the actuator's output force and frequency response based on the amplitude and frequency characteristics of the residual vibration signal, thereby maximizing the suppression of seat vibration.

[0148] Through the above technical solution, this application enables real-time and precise control of vehicle seat vibration. By employing a periodically corrected reverse damping signal generation method, the system can quickly adapt to dynamic changes in vibration characteristics during vehicle operation. Simultaneously, feedback adjustment of residual vibration signals further improves the stability and sustainability of the damping effect. This closed-loop control strategy effectively suppresses seat vibrations of various frequencies and amplitudes, significantly improves ride comfort, and creates a more stable signal acquisition environment for monitoring in-vehicle physiological parameters.

[0149] In some of the solutions described above in this application, due to the dynamic change characteristics of the seat vibration signal and the effect of actuator response delay, it is difficult to continuously optimize the vibration cancellation effect by using only the initially designed reverse damping signal, resulting in fluctuations in the residual vibration signal, which in turn affects the real-time performance and stability of the vibration reduction control.

[0150] In response, this application further proposes an improvement scheme: adjusting the filter coefficients based on the residual vibration signal and the convergence coefficient, and generating a reverse damping signal based on the adjusted filter coefficients and the seat vibration signal. The updating process of the filter coefficients involves the preprocessing of the transfer function and an error feedback mechanism, forming a closed-loop control system.

[0151] The adjustment of the filter coefficients is driven by the product of the error signal and the preprocessed signal. The error signal is calculated from the deviation between the actual vibration reduction effect and the expected output, while the preprocessed signal is generated by inputting the original vibration signal into the inverse model of the transfer function. The convergence coefficient is set to a range of 0.001 to 0.1 to balance the algorithm's convergence speed and system stability. The introduction of the transfer function considers the physical response characteristics between the actuator and the sensor; for example, it can be set as a second-order low-pass filter model to match the frequency characteristics of the mechanical system. The iterative update of the filter coefficients uses a least mean square algorithm, with each update determined by a linear combination of the preprocessed signal and the error signal. The generation of the reverse vibration reduction signal is achieved through the convolution operation of the real-time updated filter coefficients and the original vibration signal. The window length of the convolution operation can be set to 16 to 64 sampling points to balance computational efficiency and dynamic response capability.

[0152] Specifically, vibration sensors continuously collect seat vibration signals and input them into an adaptive filter. Simultaneously, the actual damping effect output by the actuator forms a closed loop through feedback from the residual vibration sensor. Within each sampling period, the original vibration signal is preprocessed using the inverse transfer function model, multiplied by the residual vibration error signal, and then multiplied by a convergence coefficient to form the incremental adjustment of the filter coefficients. The updated filter coefficients are convolved with the current vibration signal, and the resulting reverse damping signal is processed by the actuator's physical response model before outputting the mechanical action. This process continuously corrects the filter parameters through real-time error feedback, automatically compensating for actuator response delays and system parameter drift. For example, during vehicle acceleration, the convergence coefficient can be automatically adjusted to accelerate parameter updates, while during constant speed, the convergence coefficient is reduced to improve system stability. Through this dynamic adjustment mechanism, the amplitude of the residual vibration signal is continuously suppressed, and the vibration energy decay rate is increased by more than 30% compared to a fixed-parameter system.

[0153] As a preferred embodiment, the solution of this application is specifically implemented as follows:

[0154] During vehicle operation, vibration sensors mounted on the seats collect seat vibration signals x(n) in real time. Simultaneously, residual vibration sensors acquire the damped residual seat vibration signal e(n). Based on these two signals, the system dynamically adjusts the filter coefficient w(n) and the reverse damping signal d(n) to achieve adaptive cancellation of seat vibration.

[0155] Specifically, the filter coefficients w(n) are first adjusted based on the residual vibration signal e(n) and a preset convergence coefficient u. The convergence coefficient u can be set to a value between 0.01 and 0.1 to control the convergence speed and stability of the algorithm. The update formula for the filter coefficients w(n) is:

[0156] w(n) = w(n-1) - u*x'(n)*e(n);

[0157] Here, x'(n) is the output of the seat vibration signal x(n) after being filtered by the transfer function s'(n). The transfer function s'(n) is used to simulate the physical transmission characteristics between the actuator and the vibration signal, and can be obtained through system identification methods.

[0158] Next, based on the updated filter coefficients w(n) and the seat vibration signal x(n), the reverse damping signal d(n) is adjusted. The generation process of the reverse damping signal d(n) is as follows:

[0159] y(n) = x(n) * w(n);

[0160] d(n) = s(n) * y(n);

[0161] Where s(n) is the transfer function of the actuator, and y(n) is the output of x(n) after being filtered by w(n).

[0162] Finally, the generated reverse damping signal d(n) is applied to the seat via an actuator to counteract the original vibration. The system continuously monitors the residual vibration signal e(n) and, according to the formula:

[0163] e(n) = d(n) - s(n) * y(n);

[0164] The actual vibration reduction effect is calculated. Through this closed-loop control method, the system can continuously optimize the filter coefficients and the reverse damping signal, so that the residual vibration signal e(n) gradually decreases, thereby achieving active vibration reduction of the seat vibration.

[0165] Through the above technical solutions, this application effectively addresses the problems caused by dynamic changes in seat vibration signals and actuator response delays during vehicle operation. By employing an adaptive filtering algorithm and closed-loop control mechanism, the system can dynamically adjust vibration reduction parameters based on real-time vibration conditions, overcoming the poor adaptability of fixed-parameter vibration reduction schemes in complex vibration environments. Simultaneously, by introducing a transfer function model, the physical transmission characteristics between the actuator and the vibration signal are considered, resulting in more precise control. Furthermore, the application of the Least Mean Square (LMS) algorithm enables the system to quickly track the time-varying characteristics of the vibration signal, effectively suppressing periodic vibration interference. The synergistic effect of these technologies significantly improves the real-time performance, stability, and vibration reduction effect of the active vibration damping system, providing a more comfortable riding experience for vehicle occupants.

[0166] In some of the solutions mentioned above in this application, a collaborative control scheme for physiological parameter detection and active vibration reduction based on the user's presence status was proposed. However, in scenarios where there are multiple seats in a vehicle, if physiological parameter acquisition and active vibration reduction functions are activated for all seats at the same time, it will lead to a waste of system resources, an increase in data processing burden, and may cause invalid signal interference due to false triggering of sensors in unoccupied seats, affecting the overall system's operating efficiency and accuracy.

[0167] In response, this application further proposes a judgment mechanism: when the vehicle is in motion and a user is in the vehicle seat, determine physiological parameters or activate the active damping function.

[0168] The user's presence is determined using a pressure sensor array or an infrared thermal imaging module. The pressure sensors are distributed in a grid pattern on the seat surface; a user presence is detected when a continuous pressure distribution matches the characteristics of a human sitting posture. Vehicle driving status is acquired by real-time reading of the vehicle speed signal via the onboard bus; a driving state is determined when the vehicle speed consistently exceeds a set threshold. The seat-level function control module establishes independent communication links with the sensor groups of each seat. When a specific seat meets dual conditions, an activation command is sent only to the physiological parameter acquisition module of that seat, while a reverse vibration signal generation command is sent to the actuator of that seat. After the physiological parameter acquisition module is activated, the cardiac impact signal sensor acquires the raw signal at a sampling rate of no less than 100Hz, while the vibration damping control signal received by the actuator is delayed by no more than 5ms, ensuring real-time vibration suppression. The seat status detection cycle can be set to execute once per second. When a user is detected leaving the seat, the physiological parameter acquisition and actuator power are cut off within 200ms to avoid noise generated by sensor idle operation.

[0169] Specifically, when the vehicle is in motion, the pressure sensors of each seat continuously monitor the pressure distribution on the contact surface. When a seat detects a pressure center offset of less than 10mm and a total pressure value exceeding 40kg, it is determined that a user is present in that seat. At this time, the vehicle control unit allocates an independent signal processing channel to that seat, activates the cardiac impact signal sensor and residual vibration sensor, and simultaneously initiates the adaptive filtering algorithm of the actuator under the seat. The actuator generates a reverse vibration waveform with a 180-degree phase difference based on the raw vibration signal collected by the vibration sensor, and applies a mechanical reaction force through the electromagnetic actuator. The residual vibration sensor monitors the vibration energy of the seat after vibration reduction in real time. When the vibration energy exceeds a preset threshold, the convergence coefficient of the adaptive filter is automatically adjusted, changing the filter coefficient update step size from 0.01 to 0.05 to accelerate the vibration suppression process. After activation, the physiological parameter acquisition module automatically blocks the sensor data input from other unused seats, processing only the mixed signal from the current seat. When the vehicle stops or the user leaves the seat for more than 30 seconds, the system automatically releases the computing resources occupied by that seat and allocates them to the processing channels of other seats currently in use.

[0170] As a preferred embodiment, the solution of this application is implemented as follows: The vehicle main controller detects the pressure distribution of each occupant area in real time through a seat pressure sensor array. When the pressure value of the right-side seat in the second row exceeds 20 kg for 5 consecutive seconds, it is determined that the seat is occupied. At this time, the vehicle speed sensor detects that the current vehicle speed is 60 km / h, and the ECU determines that the vehicle is in motion. The control module performs a logical AND operation between the seat occupancy status and the vehicle's driving status to generate a function activation command, supplying power only to the impact signal acquisition sensor of the right-side seat in the second row, and simultaneously activating the linear actuator under the seat. The actuator receives vibration signals in the 20-200 Hz frequency band from the vibration sensor, generates a reverse drive current through a phase reversal algorithm, and causes the vibration damping actuator to produce mechanical movement to counteract the vibration. During this period, the unoccupied left-side seat in the first row keeps the sensor de-energized, and its corresponding actuator circuit is in a high-impedance state.

[0171] Through the above technical solution, this application achieves precise energy consumption control in multi-seat vehicle scenarios, centrally allocating system resources to the seats actually in use, reducing the number of data processing channels by more than 50%. This solution effectively avoids white noise interference from idle seat sensors affecting the core processor, reducing the false trigger rate to below 0.5%. Through a dynamic power management mechanism, the overall system power consumption is reduced by approximately 40%, while ensuring that the signal-to-noise ratio of physiological parameter acquisition signals from the active seats is improved by more than 15dB.

[0172] In some of the solutions mentioned above in this application, a scheme to improve the stability of physiological parameter detection through multi-sensor fusion was proposed. However, during vehicle operation, due to factors such as changes in user posture and fluctuations in sensor contact state, the signal quality of different sensors will change dynamically. If a fixed signal source or a simple weighting method is used, it will cause the physiological parameter detection results to jump or decrease in reliability. In particular, when the sensor signal quality suddenly deteriorates, the existing solutions lack a dynamic weight adjustment mechanism, making it difficult to maintain the continuity of detection results.

[0173] To address this, this application further proposes an improved solution: determining the physiological parameter signals based on the signal quality of each physiological parameter signal acquisition sensor and the acquired signals of each physiological parameter signal acquisition sensor. This further includes: determining the weight of each physiological parameter signal acquisition sensor based on its signal quality; and processing the acquired signals of each physiological parameter signal acquisition sensor according to its weight to obtain the physiological parameter signals.

[0174] The signal quality dynamic monitoring system allows for setting quality levels based on signal-to-noise ratio (SNR) thresholds. For example, a SNR higher than 30 dB is defined as a high-quality signal source. Weight allocation employs a linear or non-linear mapping relationship, with the weight increasing by 20% for every 5 dB increase in SNR. The weight adjustment module periodically calculates the standard deviation of physiological parameters. When the standard deviation exceeds 0.5 bpm, weight redistribution is triggered to reduce the contribution of the fluctuation sensor. Smooth transition utilizes a linear interpolation algorithm with a 5-second time window. When attitude changes trigger signal source switching, the weight of the new signal source linearly increases from 0% to 100%, while the weight of the old signal source decreases synchronously, ensuring that at least two signal sources participate in fusion during the transition.

[0175] Specifically, multiple physiological signals are acquired in real time through a vibration sensor array. The signal-to-noise ratio (SNR) of each signal is calculated using a Fast Fourier Transform (FFT) to determine the ratio of the energy in the main frequency band to the energy in the noise band. When the SNR of a signal drops below 25 dB, its weight coefficient decreases exponentially, for example, by 0.8 per second, while the weights of adjacent sensors are increased. Physiological parameter stability is calculated using a sliding window variance with a window length of 10 seconds. When the variance exceeds a threshold, a weight redistribution process is initiated. When the seat back angle changes by more than 15 degrees, a posture change flag is triggered. At this time, the weight of the newly activated backrest sensor gradually increases from its initial value to its maximum value within 5 seconds, while the weight of the original seat cushion sensor decreases synchronously. Kalman filtering is used to achieve seamless integration of multi-source data. The resulting closed-loop control system can dynamically optimize the signal source combination, maintain the continuity of heart rate detection results under bumpy vehicle conditions, and control the data fluctuation amplitude within ±2 bpm.

[0176] As a preferred embodiment, the solution of this application is implemented as follows: During vehicle operation, multiple physiological parameter signal acquisition sensors built into the seat collect the user's heart rate signal in real time. The signal strength and signal-to-noise ratio (SNR) parameters output by each sensor are monitored. The signal strength is obtained by calculating the root mean square value of the signal amplitude within the effective frequency band, and the SNR is calculated by comparing the ratio of the power in the effective frequency band to the power in the noise frequency band. Each sensor is scored for signal quality every 2 seconds. The signal quality score is positively correlated with the weight, and the weight calculation formula uses a normalized exponential function. The sensor with the highest current weight is selected as the main signal source, while the signal from the sensor with the second highest weight is retained as redundancy. When a change in user posture is detected that causes a drop in the SNR of the main signal source exceeding 30%, a signal source switching mechanism is triggered. During the switching process, the weights of the main and secondary signal sources transition according to an exponential decay law, with a transition period set to 3 seconds. During the transition, the data from the main and secondary signal sources are output after dynamic updates of the fusion weights. Furthermore, the system continuously tracks the standard deviation of the final heart rate value. If the standard deviation exceeds a preset threshold for five consecutive cycles, the weight allocation ratio of the current main signal source is automatically reduced, and the fusion weights of each sensor are recalculated.

[0177] Through the above technical solutions, this application solves the technical problem of data jumps during multi-sensor signal source switching. The dynamic weight allocation mechanism enables the system to select the optimal signal source in real time, avoiding detection failure due to signal degradation of a single sensor; the closed-loop stability feedback adjustment effectively suppresses weight allocation deviation, ensuring the continuity and reliability of output parameters; the smooth transition signal switching method eliminates abrupt changes in heart rate data under sudden attitude change scenarios, especially when vehicle bumps cause frequent fluctuations in sensor contact status, maintaining the smoothness of physiological parameter output.

[0178] In some of the solutions described above in this application, the weights of the sensor are dynamically adjusted and the signal source is selected based on the signal quality of the physiological parameter signal acquisition to improve the stability of heart rate detection. However, if the signal source is switched or the weights are adjusted directly when the user's posture changes, it will cause a sudden change in the physiological parameter signal, affecting the continuity and accuracy of the detection results.

[0179] In response, this application further proposes a signal source switching mechanism: when a change in posture is detected, a smooth transition method is used to switch the signal source to avoid sudden changes in heart rate data.

[0180] The detection of posture changes can be achieved through pressure distribution sensors or inertial measurement units. For example, a posture change event is triggered when the pressure distribution on the seat surface changes beyond a preset threshold. The smooth transition method includes two implementation forms: the first is to linearly adjust the signal source weights within a preset time window, for example, gradually reducing the original signal source weight from 100% to 0% within 5 seconds, while simultaneously gradually increasing the new signal source weight from 0% to 100%; the second is to fuse multi-source signals using a moving average algorithm, for example, mixing the old and new signal source data in an exponentially decaying weighted manner, with the mixing ratio dynamically changing over time. During signal source switching, the heart rate variability coefficient within the transition interval needs to be monitored synchronously. When the variability coefficient exceeds a safety threshold, the transition time window is automatically extended, for example, extending the default 5-second transition period to 8 seconds.

[0181] Specifically, when the seat pressure sensor detects a pressure center offset exceeding 30mm, it is determined to be a posture change event. At this point, a dual-signal-source parallel acquisition mode is activated, with the main control unit simultaneously receiving data streams from both the chest strap heart rate sensor and the seat's built-in optical sensor. During the first 3 seconds of the transition phase, the system uses the chest strap sensor data as a baseline to calibrate the optical sensor data, eliminating baseline drift caused by changes in contact pressure. Subsequently, a weight adjustment phase begins, where the system decreases the weight of the chest strap sensor by 20% per second while simultaneously increasing the weight of the optical sensor. During this process, Kalman filtering is used to dynamically fuse the two signals. If a difference in the interval between adjacent heartbeats exceeds 200ms during the transition, an automatic transition period extension mechanism is triggered, adjusting the total transition time from 5 seconds to 7 seconds. An abnormal data is then smoothed using a cubic spline interpolation algorithm. Finally, the output heart rate data, after being filtered by moving average, has its instantaneous fluctuation controlled within ±2bpm, effectively avoiding data jumps caused by signal source switching.

[0182] As a preferred embodiment, the solution of this application is implemented as follows: When the accelerometer built into the seat detects a change in the user's posture angle exceeding 15 degrees, a posture change event is triggered. Within a 5-second time window after the posture switch is triggered, the system linearly decreases the weight coefficient of the original signal source from its initial value of 1.0 to 0.2, while simultaneously linearly increasing the weight coefficient of the new signal source from 0.0 to 0.8. Each weight update cycle is 100 milliseconds, and each weight adjustment step size is 0.08. The heart rate data from the old and new signal sources are mixed in the fusion module using a moving average algorithm. The mixed output value is the weighted sum of the original signal source value multiplied by its current weight and the new signal source value multiplied by its current weight. When the time window ends, if the signal-to-noise ratio of the new signal source remains above 35dB, the system completely switches to the new signal source.

[0183] Through the above technical solution, this application effectively eliminates the data step phenomenon generated during signal source switching, ensuring that the heart rate waveform curve maintains a continuous and smooth transition. In scenarios where the user's body tilts due to vehicle turning, the system can maintain heart rate fluctuations within ±3 bpm, avoiding false alarms caused by sudden changes in posture. Simultaneously, this solution reduces the parameter reset frequency of the subsequent Kalman filter, shortening the convergence time of the heart rate fusion algorithm by 40%, significantly improving monitoring stability in dynamic environments.

[0184] The following section uses a vehicle equipped with heart rate monitoring as an example to illustrate the physiological parameter monitoring method provided in this application.

[0185] This application provides a method and system for calculating heart rate under vehicle driving conditions based on active vibration reduction and adaptive filtering, which can effectively suppress the interference of vehicle body vibration on heart rate measurement sensors and improve the accuracy of heart rate measurement.

[0186] This application collects the user's cardiac impact signal by placing a non-contact sensor (which does not need to be attached to the user's body) on the car seat. The collected signal is then bandpass filtered from 0.5Hz to 5Hz to further improve the signal-to-noise ratio of the collected cardiac impact signal. Next, peak detection and power spectrum estimation are performed on the filtered cardiac impact signal. The detected peak values ​​are smoothed and subtracted. The peak-to-peak interval (RR) is calculated by differentiating the smoothed peak values. According to the time-domain heart rate calculation formula HR = 60 / RR, the time-domain candidate heart rate HRt is obtained. To prevent errors in peak-to-peak detection, peak detection is performed on the calculated cardiac impact signal power spectrum. According to the frequency-domain heart rate calculation formula HR = 60*F, the frequency-domain candidate heart rate HRf is obtained. HRt and HRf are then passed to the heart rate confirmation module, which completes the final heart rate calculation.

[0187] Specifically, such as Figure 2 As shown, this application provides a physiological parameter monitoring system for improving the signal-to-noise ratio of cardiac impact signals; the system includes a monitoring module 201, an active damping module 202, a signal-to-noise ratio enhancement module 203, and a heart rate measurement module 204, wherein:

[0188] The monitoring module 201 is used to monitor whether the vehicle is moving and whether there is a user above the seat. It can be installed outside or inside the seat and is connected to the active damping module 202 via wired or wireless means.

[0189] The active damping module 202 can be installed in one or more vehicle seats. When the vehicle is in motion, the active damping module located inside the seat is activated. The active damping module collects and analyzes the frequency components of the seat vibration signal in real time based on the vibration sensor, and then sends the seat vibration signal components to the controller of the active damping module. The controller generates a vibration signal with the same amplitude as the seat vibration amplitude but opposite phase. The generated vibration signal is sent to the actuator of the active damping module to act on the seat to counteract the influence of the seat vibration on the heart rate acquisition sensor.

[0190] The signal-to-noise ratio (SNR) enhancement module 203 can be composed of two cushions and an adaptive filtering algorithm for cardiac impact signals. One cushion is a flexible damping cushion that can reduce vibration, and the other is a flexible cushion with vibration signal acquisition sensors placed on top and bottom. The seat vibration signal caused by vehicle movement is processed by the active damping device and further weakened by the flexible damping cushion. Therefore, the interference of seat vibration on the heart rate acquisition sensor can be further reduced. Adaptive filtering algorithm for cardiac impact signals: When the user sits on the flexible cushion located in the SNR enhancement module, the vibration sensor under the cushion collects the residual seat vibration signal after processing by the active damping device and the damping cushion, and uses it as the noise reference signal for the adaptive filtering algorithm module for cardiac impact signals. The vibration sensor above the cushion collects the user's cardiac impact signal and uses it as the input signal for the adaptive filtering algorithm module for cardiac impact signals. Then, through the adaptive filtering algorithm, the cardiac impact signal and seat vibration noise signal are effectively separated, improving the SNR of the cardiac impact signal.

[0191] The heart rate measurement module 204 is used to measure heart rate according to a specific algorithm. It can be set inside or outside the seat and is connected to the signal-to-noise ratio enhancement module 203 by wire or wireless means, as will be described below.

[0192] This application, by combining active vibration reduction technology and adaptive filtering algorithms, can significantly reduce the interference of seat vibration on the heart rate acquisition sensor during vehicle operation, thereby improving the accuracy of heart rate measurement while the vehicle is in motion. Specifically, it includes the following beneficial effects:

[0193] 1. It adopts cardiac impact signal heart rate measurement, which is contactless for users and provides a good user driving experience.

[0194] 2. The active vibration damping device improves the signal-to-noise ratio of cardiac impact signals during vehicle operation, thereby increasing the accuracy of heart rate readings while the vehicle is in motion.

[0195] 3. The adaptive filtering algorithm is simple to calculate, has good real-time performance, low hardware requirements, and controllable cost.

[0196] When a vehicle is in motion, the vibration of the vehicle body causes the seat to vibrate. This seat vibration not only resonates with the user's organs and tissues, affecting the user's health, but also significantly interferes with non-contact heart rate measurement. This invention primarily addresses the problem of interference with cardiac impact signals when the vehicle is in motion.

[0197] The present application will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0198] like Figure 2As shown, the system provided in this application includes: a monitoring module 201 for determining whether the vehicle is moving and whether a user is seated; an active vibration damping module 202 for reducing seat vibration when the vehicle is moving; a signal-to-noise ratio enhancement module 203 for further improving the signal-to-noise ratio of cardiac impact signals; and a heart rate measurement module 204. These modules interact with each other, enabling the active vibration damping device, signal-to-noise ratio enhancement module, and heart rate estimation module to be automatically turned off when the vehicle is stationary or no user is seated, and to be automatically turned on when the vehicle is moving and a user is seated, without requiring the user to issue commands or operate buttons, thus improving the driving experience.

[0199] Unlike passive vibration damping, the active vibration damping device of this application adopts the principle of vibration wave superposition. It uses the vibration generated by the secondary vibration source to counteract the vibration on the foundation structure under the seat, preventing the vibration from being transmitted to the seat. Its principle is as follows: Figure 5 As shown, this will be described later.

[0200] Figure 3 This is a schematic diagram of one type of seat provided in this application. Figure 4 This is a schematic diagram illustrating the setup of a specific sensor within the physiological parameter monitoring system provided in this application, such as... Figure 3 as well as Figure 4 As shown, the seat includes:

[0201] Seat base 205 is used to secure the seat inside the vehicle.

[0202] Active damping module 202, integrated into such Figure 3 The seat interior includes a vibration sensor 2021, an actuator 2022, and a controller (not shown). When the vehicle is in motion, the active damping module 202 is activated. First, the vibration sensor 2021 collects vibration signals from the seat base 205. The controller of the active damping module analyzes the frequency components of the seat vibration signals and generates signals with the same amplitude but opposite phase. The controller sends the generated signal to the actuator 2022, which acts on the seat to achieve vibration damping. The controller adjusts the operation of the actuator 2022 based on the residual vibration signal strength monitored by the residual vibration acquisition sensor 2032. The damping effect after the active damping module is controlled is as follows: Figure 5 As shown.

[0203] Seat foam 206 is used to improve seat comfort and further reduce residual vibration.

[0204] Signal-to-noise ratio enhancement module 203, integrated into such Figure 3The seat interior includes a passive damping pad 2031 (which can be a flexible damping pad for vibration reduction and damping), a residual vibration sensor 2032, a signal isolation pad 2033 (which can be a flexible pad with vibration signal acquisition sensors arranged vertically), and a cardiac impact signal acquisition sensor 2034. The cardiac impact signal acquisition sensor 2034 is located above the signal isolation pad 2033, and the residual vibration sensor 2032 is located below it. The seat vibration signal (i.e., the residual vibration signal) after being damped by the active damping module is transmitted to the passive damping pad via the seat foam 206, further weakening the seat vibration signal. Then, the seat vibration signal is transmitted to the residual vibration sensor 2032 below the signal isolation pad 2033.

[0205] like Figure 6 This is a schematic diagram of the active vibration damping control algorithm corresponding to the active vibration damping module 202. Here, x(n) represents the seat vibration signal collected by the vibration sensor, d(n) represents the vibration source signal, w(n) is the filter coefficient, y(n) is the output of x(n) after filtering by w(n), LMS is the control algorithm, s(n) is the transfer function, and e(n) is the residual vibration signal of the seat after vibration damping. Its calculation formula is as follows:

[0206] e(n) = d(n) - s(n) * y(n);

[0207] The update formula for y(n) is as follows:

[0208] y(n) = x(n) * w(n);

[0209] The formula for updating the filter coefficients w(n) is as follows:

[0210] x'(n) = x(n) * s'(n);

[0211] w(n) = w(n-1) - u*x'(n)*e(n);

[0212] Where x'(n) is the output of the x(n) signal after filtering by the transfer function s'(n), s(n) is the transfer function, s'(n) is an estimate of s(n), and u is the convergence coefficient; based on the adjustment of the filter coefficient w(n) according to e(n) and the convergence coefficient u, the residual vibration signal e(n) of the seat after vibration reduction continuously decreases, thereby realizing active vibration reduction of the seat; the vibration reduction effect after the active vibration reduction module is controlled is as follows: Figure 5 As shown, the red signal is the residual signal e(n) after shock absorption, and the vibration amplitude is greatly reduced compared with the seat vibration signal x(n).

[0213] Before the signal-to-noise ratio enhancement module 203 sends the signal n2 collected by the residual vibration acquisition sensor 2032 and the signal (n1+s1) collected by the cardiac impact signal acquisition sensor 2034 to the heart rate measurement module 204, it performs the following steps: Figure 7 The adaptive filtering algorithm shown is as follows: s1 is the effective cardiac impact signal acquired by the cardiac impact signal acquisition sensor; n1 is the seat vibration signal acquired by the cardiac impact signal acquisition sensor; n2 is the seat vibration signal acquired by the residual seat vibration sensor; the cardiac impact signal ￠(n) obtained by the algorithm corresponding to the adaptive filter is calculated as follows:

[0214] ￠(n)=s1+n1-y;

[0215] Where y is the output of n² after passing through filter W; its calculation formula is as follows:

[0216] y = n² * w;

[0217] Furthermore, the update formula for filter W is as follows:

[0218] W = Wp - c * ￠(n) * n²;

[0219] Where c is the convergence coefficient of the adaptive filter; Wp is the filter coefficient at the previous time step.

[0220] like Figure 8 As shown, the adaptive filtering algorithm provides the residual vibration signal component of the seat based on the signal collected by the seat residual vibration sensor, and further effectively removes the seat vibration signal collected by the cardiac impact signal, thereby improving the signal-to-noise ratio of the cardiac impact signal.

[0221] like Figure 9 As shown, after active vibration reduction, passive vibration reduction, and adaptive filtering, the processed signal is sent to the heart rate measurement module 204. The heart rate measurement module 204 performs the following steps:

[0222] S901: Bandpass filtering is performed on the input cardiac impulse signal.

[0223] This step performs bandpass filtering on the input cardiac impulse signal ￠(n), with a bandwidth of 0.5Hz to 5Hz. The specific implementation method has been described above and will not be repeated here. Then, the filtered cardiac impulse signal is processed in two ways.

[0224] S902: Time-domain peak-to-peak value detection is performed on the filtered cardiac impulse signal.

[0225] The first process involves time-domain peak-to-peak detection of the filtered cardiac impulse signal, i.e., peak and trough detection of the filtered signal. This is done by searching for the maximum value of the acquired signal within a certain time period and saving the time tx when the maximum value occurs. tx is then saved to a buffer. When the number of results saved in the buffer is greater than or equal to N, the peak-to-peak interval PS of the cardiac impulse signal is obtained by differentiating the results in the buffer. The peak-to-peak interval sequence PS is then smoothed and filtered to remove intervals less than 0.8 times the mean and more than 1.2 times the mean, resulting in the filtered peak-to-peak interval sequence PSF. The median RR_med of the latest M peak-to-peak intervals in the PSF sequence is calculated. Using the heart rate calculation formula HR = 60 / RR_med, the candidate heart rate HR1 is obtained, which is the time-domain candidate parameter mentioned above.

[0226] S903: Perform power spectrum analysis on the filtered cardiac impulse signal.

[0227] The second process involves power spectrum analysis of the filtered cardiac impact signal. Specifically, FFT transformation is performed on the cardiac impact signal acquired for a time exceeding tn to obtain its power spectrum. Spectral peaks in the power spectrum range of 0.7Hz to 2.5Hz are searched, and the obtained spectral peak hf is substituted into the heart rate calculation formula Hr = 60 * hf to obtain the candidate heart rate HR2, which is the frequency domain candidate parameter mentioned above.

[0228] S904: Perform candidate parameter fusion to obtain the final heart rate.

[0229] This step performs Kalman fusion on candidate heart rates HR1 and HR2 to obtain the final heart rate HR_f, which is the physiological parameter mentioned above.

[0230] In summary, this embodiment, by combining active vibration reduction technology and adaptive filtering algorithm, can significantly reduce the interference of seat vibration on the heart rate acquisition sensor when the vehicle is in motion, and improve the accuracy of heart rate measurement while the vehicle is in motion.

[0231] According to a second aspect of this application, an embodiment of this application also provides a physiological parameter monitoring device, which includes: a monitoring module, used to determine physiological parameters based on residual vibration signals collected by residual vibration acquisition sensors in the vehicle seat and physiological parameter signals collected by physiological parameter signal acquisition sensors in the vehicle seat when the vehicle is in motion.

[0232] According to a third aspect of this application, embodiments of this application also provide a non-transitory computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the steps of the above-described physiological parameter monitoring method. This non-transitory computer-readable storage medium possesses all the beneficial effects of the above-described physiological parameter monitoring method, which will not be elaborated further here.

[0233] According to a fourth aspect of this application, embodiments of this application also provide a computer program product, including a computer program that, when executed by a processor, implements the steps of the above-described physiological parameter monitoring method. This computer program product possesses all the beneficial effects of the above-described physiological parameter monitoring method, which will not be elaborated upon further herein.

[0234] According to a fifth aspect of this application, embodiments of this application also provide an electronic device, including: a memory and a processor, wherein a computer program is stored in the memory; the processor is used to execute the computer program in the memory to implement the steps of the above-described physiological parameter monitoring method. This electronic device possesses all the beneficial effects of the above-described physiological parameter monitoring method, which will not be elaborated further here.

[0235] Computer-readable storage media can be, for example, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatuses, or devices, or any combination thereof, without particular limitation herein. More specific examples of computer-readable storage media may include, but are not limited to: electrical connections having one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0236] In some embodiments of this application, a computer-readable storage medium may be any tangible medium that contains or stores a program that may be used or combined with an instruction execution system, apparatus, or device.

[0237] The aforementioned computer-readable storage medium may be included in the aforementioned electronic device, or it may exist independently and not assembled into the electronic device. The aforementioned computer-readable storage medium carries one or more programs that, when executed by the electronic device, cause the electronic device to:

[0238] While the vehicle is in motion, physiological parameters are determined based on residual vibration signals collected by residual vibration acquisition sensors in the vehicle seat and physiological parameter signals collected by physiological parameter signal acquisition sensors in the vehicle seat.

[0239] Computer program code for performing operations of some embodiments of this application can be written in one or more programming languages ​​or a combination thereof. These programming languages ​​include object-oriented programming languages ​​such as Java, Smalltalk, and C++, as well as conventional procedural programming languages ​​such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network (including a local area network (LAN) or a wide area network (WAN)), or it can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0240] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function.

[0241] It should also be noted that in some alternative implementations, the functions marked in the box may occur in a different order than those marked in the attached figures.

[0242] For example, two consecutively represented blocks can actually be executed in substantially parallel order, and sometimes they can be executed in reverse order, depending on the functions involved. It should also be noted that each block in a block diagram and / or flowchart, as well as combinations of blocks in block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified functions or operations, or using a combination of dedicated hardware and computer instructions.

[0243] The units described in some embodiments of this application can be implemented in software or in hardware. The described units can also be located in a processor.

[0244] The functions described above in this document can be performed, at least in part, by one or more hardware logic components. For example, exemplary types of hardware logic components that can be used, without limitation, include: Field Programmable Gate Arrays (FPGAs), Application-Specific Integrated Circuits (ASICs), Application Standard Products (ASSPs), System-on-Chip (SoCs), Complex Programmable Logic Devices (CPLDs), and so on.

[0245] According to a sixth aspect of this application, embodiments of this application also provide a physiological parameter monitoring system, comprising:

[0246] A signal-to-noise ratio (SNR) enhancement device, installed inside a vehicle seat, includes a signal isolation pad, a residual vibration acquisition sensor located on one side of the signal isolation pad, and a physiological parameter signal acquisition sensor located on the other side of the signal isolation pad; and

[0247] Physiological parameter measuring device;

[0248] in:

[0249] The residual vibration acquisition sensor is configured to acquire residual vibration signals when the vehicle is in motion.

[0250] The physiological parameter signal acquisition sensor is configured to acquire physiological parameter signals when the vehicle is in motion.

[0251] The physiological parameter measuring device is configured to determine physiological parameters based on the residual vibration signal and the physiological parameter signal.

[0252] Optionally, the aforementioned system further includes:

[0253] An active damping device, installed within the vehicle seat, includes a vibration sensor, a vibration source, and an actuator; wherein:

[0254] The vibration sensor is configured to collect seat vibration signals when the vehicle is in motion.

[0255] The vibration source is configured to control the actuator to perform active vibration reduction based on the seat vibration signal and the residual vibration signal.

[0256] Optionally, the aforementioned system further includes:

[0257] The monitoring module is configured to enable the physiological parameter monitoring system when the vehicle is in motion and a user is present in the vehicle seat.

[0258] According to the seventh aspect of this application, such as Figure 10 As shown in the illustration, this application also provides a vehicle 10, which includes the aforementioned electronic equipment. This vehicle possesses all the beneficial effects of the aforementioned electronic equipment, etc., which will not be elaborated upon further herein.

[0259] The vehicle may be a gasoline-powered vehicle, a plug-in hybrid electric vehicle, or a new energy vehicle, etc., and this application does not make any specific restrictions.

[0260] In the description of this application, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include one or more features. In the description of this application, "multiple" means two or more, unless otherwise explicitly specified.

[0261] In the above embodiments, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.

[0262] The embodiments, implementation methods, and related technical features of this application can be combined and substituted for each other without conflict.

[0263] The above are merely preferred embodiments of this application and are not intended to limit this application in any way. Although the descriptions of each embodiment in this application have different focuses, and parts not described in detail in a certain embodiment can be referred to the relevant descriptions of other embodiments, any simple modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of this application without departing from the content of the technical solution of this application shall still fall within the scope of the technical solution of this application.

Claims

1. A method for monitoring physiological parameters, characterized in that, include: While the vehicle is in motion, physiological parameters are determined based on residual vibration signals collected by residual vibration acquisition sensors in the vehicle seat and physiological parameter signals collected by physiological parameter signal acquisition sensors in the vehicle seat.

2. The method according to claim 1, characterized in that, The determination of physiological parameters based on residual vibration signals collected by residual vibration sensors in the vehicle seat and physiological parameter signals collected by physiological parameter signal sensors in the vehicle seat includes: Based on the residual vibration signal, an effective physiological signal is obtained from the physiological parameter signal; The physiological parameters are determined based on the effective physiological signals.

3. The method according to claim 2, characterized in that, The step of separating the effective physiological signal from the physiological parameter signal based on the residual vibration signal includes: An adaptive filtering method is used to filter out the associated signals of the residual vibration signals from the physiological parameter signals based on the residual vibration signals, so as to obtain the effective physiological signals.

4. The method according to claim 2, characterized in that, Determining the physiological parameters based on the effective physiological signals includes: The effective physiological signal is subjected to bandpass filtering to obtain the filtered signal; Based on the filtered signal, time-domain candidate parameters and frequency-domain candidate parameters are obtained; The physiological parameters are generated based on the time-domain candidate parameters and the frequency-domain candidate parameters.

5. The method according to claim 4, characterized in that, The frequency of the bandpass filter is from 0.5 Hz to 5 Hz.

6. The method according to claim 4, characterized in that, The process of obtaining time-domain candidate parameters and frequency-domain candidate parameters based on the filtered signal includes: Peak and trough detection is performed on the filtered signal to obtain the median of the peak-to-peak interval of the filtered signal; The time-domain candidate parameters are determined based on the median. The power spectrum of the filtered signal is obtained by transforming the filtered signal. The frequency domain candidate parameters are determined based on the spectral peaks of the power spectrum.

7. The method according to claim 4, characterized in that, The step of generating the physiological parameters based on the time-domain candidate parameters and the frequency-domain candidate parameters includes: The physiological parameters are obtained by Kalman fusion of the time-domain candidate parameters and the frequency-domain candidate parameters.

8. The method according to claim 1, characterized in that, Also includes: When the vehicle is in motion, the vehicle seat is controlled to actively dampen vibrations based on the seat vibration signals collected by the vibration sensors in the vehicle seat and the residual vibration signals.

9. The method according to claim 8, characterized in that, The step of controlling the vehicle seat to actively dampen vibrations based on the seat vibration signals collected by the vibration sensors in the vehicle seat and the residual vibration signals includes: A reverse damping signal is generated based on the seat vibration signal; The actuator in the vehicle seat is controlled to operate according to the reverse damping signal; The actuator is adjusted based on the residual vibration signal.

10. The method according to claim 9, characterized in that, Adjusting the actuator's operation based on the residual vibration signal includes: Based on the residual vibration signal and the convergence coefficient, adjust the filter coefficients; The reverse damping signal is adjusted based on the filter coefficients and the seat vibration signal.

11. The method according to claim 1, characterized in that, Also includes: Physiological parameters are determined while the vehicle is in motion and a user is in the vehicle seat.

12. The method according to any one of claims 1 to 11, characterized in that, Also includes: The physiological parameter signals are determined based on the signal quality of each physiological parameter signal acquisition sensor and the acquired signals of each physiological parameter signal acquisition sensor.

13. The method according to claim 12, characterized in that, The step of determining the physiological parameter signals based on the signal quality of the sensors acquiring each physiological parameter signal and the acquired signals of the sensors includes: The weight of each physiological parameter signal acquisition sensor is determined based on the signal quality of each sensor. The physiological parameter signals are obtained by processing the acquired signals from each physiological parameter signal acquisition sensor according to the weight of each sensor.

14. The method according to claim 12, characterized in that, Also includes: When a change in attitude is detected, a smooth transition method is used to switch the signal source.

15. A physiological parameter monitoring device, characterized in that, include: The monitoring module is used to determine physiological parameters based on residual vibration signals collected by residual vibration acquisition sensors in the vehicle seat and physiological parameter signals collected by physiological parameter signal acquisition sensors in the vehicle seat when the vehicle is in motion.

16. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the physiological parameter monitoring method according to any one of claims 1 to 14.

17. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements the physiological parameter monitoring method according to any one of claims 1 to 14.

18. An electronic device, characterized in that, include: A memory on which computer programs are stored; A processor for executing the computer program in the memory to implement the physiological parameter monitoring method according to any one of claims 1 to 14.

19. A physiological parameter monitoring system, characterized in that, include: A signal-to-noise ratio (SNR) enhancement device is installed inside a vehicle seat. The SNR enhancement device includes a signal isolation pad, a residual vibration acquisition sensor installed on one side of the signal isolation pad, and a physiological parameter signal acquisition sensor installed on the other side of the signal isolation pad. as well as Physiological parameter measuring device; in: The residual vibration acquisition sensor is configured to acquire residual vibration signals when the vehicle is in motion. The physiological parameter signal acquisition sensor is configured to acquire physiological parameter signals when the vehicle is in motion. The physiological parameter measuring device is configured to determine physiological parameters based on the residual vibration signal and the physiological parameter signal.

20. The system according to claim 19, characterized in that, Also includes: An active damping device, installed within the vehicle seat, includes a vibration sensor, a vibration source, and an actuator; wherein: The vibration sensor is configured to collect seat vibration signals when the vehicle is in motion. The vibration source is configured to control the actuator to perform active vibration reduction based on the seat vibration signal and the residual vibration signal.

21. The system according to claim 19 or 20, characterized in that, Also includes: The monitoring module is configured to enable the physiological parameter monitoring system when the vehicle is in motion and a user is present in the vehicle seat.

22. A vehicle, characterized in that, This includes the electronic device as described in claim 18, or the physiological parameter monitoring system as described in any one of claims 19 to 21.

Images