Directional noise reduction method for vehicle-mounted capacitive physiological signal monitoring

Abstract

The invention relates to the technical field of vehicle-mounted health monitoring, and provides a directional noise reduction method for vehicle-mounted capacitive physiological signal monitoring. The method comprises the following steps: acquiring an original physiological signal of a driver and vehicle state data including an engine rotating speed and a vehicle speed in real time; calculating the main characteristic frequency of the current vehicle vibration based on the vehicle state data and generating a corresponding noise reference signal; taking the characteristic frequency as a center, and adopting a programmable comb filter to carry out preliminary noise reduction on the original physiological signal; and based on the noise reference signal, carrying out fine processing on the preliminary noise reduction signal through adaptive filtering, and outputting a final noise-reduced physiological signal. According to the method, through noise source modeling and two-stage collaborative filtering, directional and efficient suppression of vehicle vibration noise is achieved, the signal-to-noise ratio and heart rate detection accuracy are effectively improved, meanwhile, the requirement for computing resources is remarkably reduced, and the self-adaptability and reliability in a complex vehicle-mounted environment are enhanced.

Description

Technical Field

[0001] This invention relates to the field of vehicle-mounted health monitoring technology, and more specifically, to a directional noise reduction method for vehicle-mounted capacitive physiological signal monitoring. Background Technology

[0002] When using capacitive sensors integrated into the steering wheel or seat to monitor the driver's electrocardiogram (ECG), respiration, and other physiological signals, vehicle vibrations are transmitted directly to the sensors through the vehicle body, introducing strong, vehicle-state-related noise into the output signal. Traditionally, fixed-frequency filters (such as fixed-bandwidth bandpass filters) are used to attempt to filter out this noise. However, because the frequency bands of vehicle vibration noise (its fundamental frequency and harmonics) significantly overlap with the frequency bands of useful physiological signals such as ECG, this filtering method, while removing noise, inevitably damages or even filters out useful waveform features (such as the R wave in ECG), leading to inaccurate or even invalid results in subsequent heart rate detection and other analyses.

[0003] Existing technologies include some general-purpose advanced signal denoising algorithms, such as wavelet transform and adaptive filtering. However, these general-purpose algorithms have significant shortcomings when applied to this specific scenario: First, their computational complexity is usually high, requiring significant hardware costs and resource overhead to achieve real-time operation on resource-constrained automotive embedded hardware platforms, hindering productization. Second, these algorithms are general in nature and do not model and optimize for noise from the specific physical source of vehicle vibration, resulting in low denoising efficiency and a lack of specificity. Finally, existing algorithms are mostly purely software-level processing, failing to effectively coordinate with the vehicle's inherent hardware data sources (such as CAN bus and IMU), and thus unable to fully utilize real-time vehicle status information to accurately locate and track noise sources.

[0004] In summary, existing technologies, whether simple fixed filters or complex general algorithms, have failed to effectively solve the vibration noise problem in vehicle-mounted capacitive physiological signal monitoring. The core challenge lies in the inability to accurately and directionally suppress time-varying vibration noise that is closely related to vehicle conditions while maintaining real-time performance and low computational cost. Therefore, there is an urgent need in this field for a noise reduction solution that is closely integrated with vehicle operating conditions, computationally efficient, and highly targeted, to ensure the accuracy and reliability of physiological signal monitoring under harsh vehicle vibration environments. Summary of the Invention

[0005] This invention addresses the technical problems existing in the prior art by providing a directional noise reduction method for vehicle-mounted capacitive physiological signal monitoring. It solves the technical problems in vehicle-mounted capacitive physiological signal monitoring, where the frequency bands of vehicle vibration noise and electrocardiogram signals overlap, causing traditional filtering methods to severely damage useful signals during noise reduction, and general noise reduction algorithms to be computationally complex and lack specificity.

[0006] According to a first aspect of the present invention, a directional noise reduction method for vehicle-mounted capacitive physiological signal monitoring is provided, comprising: S1, real-time acquisition of the driver's raw physiological signals and vehicle status data, wherein the vehicle status data includes at least engine speed and vehicle speed; S2, Based on the vehicle status data, calculate the main characteristic frequencies of the current vehicle vibration and generate the corresponding noise reference signal; S3, using the main characteristic frequency as the center frequency, perform programmable comb filtering on the original physiological signal to obtain a preliminary noise-reduced signal; S4, based on the noise reference signal, the preliminary noise reduction signal is adaptively filtered to obtain the final noise-reduced physiological signal.

[0007] Based on the above technical solution, the present invention can also be improved as follows.

[0008] Optionally, in step S1, the raw physiological signal is acquired by a capacitive sensor integrated into the steering wheel or seat, and the vehicle status data is acquired in real time through the vehicle bus system and inertial measurement unit.

[0009] Optionally, in step S2, calculating the main characteristic frequencies of the current vehicle vibration based on the vehicle state data includes: Calculate the engine vibration fundamental frequency based on the engine speed; Based on the vehicle speed, calculate the characteristic frequency of road vibration; The main characteristic frequencies include at least the fundamental frequency of engine vibration and the characteristic frequency of road surface vibration.

[0010] Optionally, in step S2, generating the corresponding noise reference signal includes: The engine vibration fundamental frequency ( ) and its harmonics, and the characteristic frequency of the road surface vibration ( The noise reference signal is generated by synthesizing its harmonics. ; Wherein, the noise reference signal Represented as the superposition of multiple sine wave components:

[0011] Where n and m are the harmonic orders, and For the amplitude of the corresponding harmonic components, and This represents the phase of the corresponding harmonic component.

[0012] Optionally, step S3 includes: Based on the main characteristic frequencies and the system sampling rate, the notch depth control parameter N of the comb filter is calculated; Configure the coefficients of the programmable comb filter according to the parameter N; The original physiological signal is filtered using a pre-configured comb filter to obtain the preliminary noise-reduced signal.

[0013] Optionally, the transfer function of the comb filter is expressed as:

[0014] in: N is the notch depth control parameter. , The system sampling rate, For the center frequency, Obtained through the aforementioned main characteristic frequencies; This indicates a signal delay of N sampling points; The pole radius is used to control the notch depth and bandwidth; The notch depth control parameter N and the pole radius Adaptive adjustments are made based on the vehicle's real-time status.

[0015] Optionally, the step of adaptively adjusting the notch depth control parameter N based on the real-time vehicle status includes: The current engine vibration fundamental frequency is dynamically calculated based on real-time vehicle status data. And based on the current engine vibration fundamental frequency Update the notch depth control parameter N to make the notch center frequency of the comb filter... The formula for real-time adjustment based on changes in vehicle operating conditions is as follows:

[0016] in, The notch depth control parameters are updated adaptively.

[0017] Optionally, the radius of the pole can be adaptively adjusted according to the real-time status of the vehicle. The steps include: Signal-to-noise ratio based on real-time estimation The pole radius is updated to dynamically optimize the notch depth and bandwidth of the comb filter based on signal quality. The formula for updating the pole radius is:

[0018] in, The updated pole radius, This is the preset maximum signal-to-noise ratio.

[0019] Optionally, step S4 includes: The initial noise reduction signal is used as the main input. The noise reference signal is used as the reference input. They are input together into the adaptive filter; The adaptive filter is based on its current weight vector. For reference input Perform weighted calculations to obtain the filter output. ; Calculate the main input With the filter output Error between The error This refers to the physiological signal after final noise reduction; Based on the error and the reference input The filter weight vector is updated using an adaptive algorithm. It is used for subsequent signal filtering.

[0020] Optionally, step S4, the step of adaptively adjusting the adaptive filter, further includes: The adaptive algorithm employs a variable step size strategy and a step size factor. The value depends on the reference input signal The instantaneous energy is adaptively adjusted, and the calculation formula is:

[0021] Where β is the convergence rate factor and α is the stabilization constant; Furthermore, the order L of the adaptive filter is based on the current noise fundamental frequency. and system sampling rate After optimization, the optimized order is expressed as:

[0022] in, This is a fine-tuning amount based on a preset order from experience.

[0023] According to a second aspect of the present invention, a directional noise reduction system for vehicle-mounted capacitive physiological signal monitoring is provided, comprising: The signal acquisition module is used to acquire the driver's raw physiological signals and vehicle status data in real time, wherein the vehicle status data includes at least engine speed and vehicle speed. A noise reference signal generation module, connected to the signal acquisition module, is used to calculate the main characteristic frequencies of the current vehicle vibration based on the vehicle state data and generate the corresponding noise reference signal. A programmable comb filter, connected to the signal acquisition module and the noise reference signal generation module, is used to filter the original physiological signal with the characteristic frequency as the center to obtain a preliminary noise-reduced signal; An adaptive filtering module is connected to the programmable comb filter and the noise reference signal generation module, respectively, and is used to filter the preliminary noise reduction signal based on the noise reference signal to output the final noise-reduced physiological signal.

[0024] According to a third aspect of the present invention, an electronic device is provided, including a memory and a processor, wherein the processor is configured to execute a computer management program stored in the memory to implement the steps of the above-described directional noise reduction method for vehicle-mounted capacitive physiological signal monitoring.

[0025] According to a fourth aspect of the present invention, a computer-readable storage medium is provided, on which a computer management program is stored, wherein when executed by a processor, the computer management program implements the steps of the above-described directional noise reduction method for vehicle-mounted capacitive physiological signal monitoring.

[0026] This invention provides a directional noise reduction method, system, electronic device, and storage medium for in-vehicle capacitive physiological signal monitoring. It accurately calculates the characteristic frequencies of vibration noise and generates digital reference signals by acquiring real-time vehicle state data (such as engine speed and vehicle speed). A programmable comb filter is then used to quickly and efficiently eliminate periodic noise strongly correlated with vibration. An adaptive filter is then used to perform fine tracking and cancellation based on this noise reference signal. This invention actively models and cancels noise from its physical source (vehicle vibration) rather than passively filtering from mixed signals, thus significantly improving the signal-to-noise ratio in complex in-vehicle environments. Furthermore, the rational division of labor in the two-stage filtering architecture significantly reduces the overhead of embedded hardware resources while ensuring high noise reduction performance, achieving a balance between accuracy and computational efficiency. Attached Figure Description

[0027] Figure 1 A flowchart of a directional noise reduction method for vehicle-mounted capacitive physiological signal monitoring provided by the present invention; Figure 2This invention provides a block diagram of a directional noise reduction system for vehicle-mounted capacitive physiological signal monitoring. Figure 3 A schematic diagram of a possible hardware structure of an electronic device provided by the present invention; Figure 4 This is a schematic diagram of the hardware structure of a possible computer-readable storage medium provided by the present invention. Detailed Implementation

[0028] The specific embodiments of the present invention will be described in further detail below with reference to the accompanying drawings and examples. The following examples are for illustrative purposes only and are not intended to limit the scope of the invention.

[0029] In this embodiment of the invention, when collecting, processing, and storing user personal information (such as images, behavioral characteristics, etc.), the implementation of the technical solution strictly adheres to the principles of legality, legitimacy, and necessity, as well as the core rule of "notification-consent." Specifically, before information collection, the system clearly informs the user of the purpose, method, scope, and usage rules of information collection through an interactive interface, and requires the user's active authorization and consent. The entire information processing process employs data encryption, access control, and other technical measures to ensure information security, and establishes mechanisms to facilitate users' exercise of their rights (such as querying, correcting, withdrawing consent, and deleting information). For exceptions stipulated by law (such as those necessary for fulfilling statutory duties or responding to public health emergencies), their application is strictly limited to the scope and limits authorized by law, ensuring that the technical solution does not contain any content that violates the law, social morality, or harms the public interest.

[0030] Figure 1 A flowchart of a directional noise reduction method for vehicle-mounted capacitive physiological signal monitoring provided by the present invention is shown below. Figure 1 As shown, the method includes steps S1 to S4: S1, real-time acquisition of the driver's raw physiological signals and vehicle status data, wherein the vehicle status data includes at least engine speed and vehicle speed.

[0031] This step is the system's fundamental data input stage. It simultaneously acquires two signals: one is raw physiological signals (such as ECG signals) containing vehicle vibration and noise, obtained through capacitive sensors integrated into the steering wheel or seat; the other is real-time vehicle status data acquired from the vehicle's CAN bus and inertial measurement unit (IMU), primarily including engine speed and vehicle speed. This step achieves synchronous acquisition of physiological signals and noise source (vehicle vibration) status information, enabling subsequent noise modeling and targeted cancellation at the source.

[0032] S2, based on the vehicle state data, calculate the main characteristic frequencies of the current vehicle vibration and generate the corresponding noise reference signal.

[0033] This step converts the physical world into a digital model. For example, using acquired engine speed and vehicle speed data, it calculates the fundamental frequency and harmonic components of the most prevalent vibration and noise using built-in physical models (such as engine ignition frequency formulas and tire rotation frequency formulas). Then, these frequency components are combined into a digital noise reference signal. This signal is essentially a predictive model of the current vehicle vibration on an electrical signal, equivalent to obtaining the noise's "fingerprint," providing a target for subsequent precise filtering. S3, using the main characteristic frequency as the center frequency, perform programmable comb filtering on the original physiological signal to obtain a preliminary noise-reduced signal; This step is the first stage of noise reduction, focusing on quickly and efficiently attenuating strong periodic noise. A comb filter is a filter that exhibits periodic notch characteristics for specific frequencies and their harmonics. This step uses the main characteristic frequency calculated in step S2 as the center frequency to dynamically configure the filter parameters, ensuring that its notch point is precisely aligned with discrete spectral noise generated by engine vibration, etc. Like a comb, the comb filter quickly filters out high-energy noise components strongly correlated with vehicle conditions from the original signal, achieving initial signal purification and reducing the burden on the next stage of noise reduction, which involves greater computational demands.

[0034] S4, based on the noise reference signal, the preliminary noise reduction signal is adaptively filtered to obtain the final noise-reduced physiological signal.

[0035] This step is the second stage of noise reduction, focusing on the fine tracking and cancellation of residual noise and time-varying components. For example, it uses the preliminary noise-reduced signal output from step S3 as the main input and the noise reference signal generated in step S2 as the reference input, both fed into an adaptive filter (such as an LMS filter). The adaptive filter continuously compares the main input and the reference input, automatically adjusting its internal parameters through an algorithm, learning from the main input and subtracting the noise components related to the reference input. Its output error signal is the final clean physiological signal after noise reduction. This step endows the system with the ability to dynamically track noise changes, achieving fine-grained and adaptive noise reduction.

[0036] Understandably, given the deficiencies in the background technology, this invention proposes a directional noise reduction method for vehicle-mounted capacitive physiological signal monitoring. This method directly acquires physical parameters of noise sources such as engine speed and vehicle speed from sources like the CAN bus, and calculates the precise characteristic frequency model of the vibration noise (i.e., the noise reference signal) based on these parameters. This transforms traditional blind signal processing into directional cancellation based on prior knowledge of the noise source. Then, a clearly defined two-stage filtering architecture is employed. First, a programmable comb filter, a highly efficient dedicated circuit / algorithm, is used to quickly and accurately eliminate the principal components of periodic noise strongly correlated with the vehicle's state, completing most of the coarse noise reduction work. The pre-purified signal is then fed into an adaptive filter for fine processing. This transforms the computationally complex adaptive filtering algorithm from processing the original signal filled with strong noise into a pre-purified signal, greatly reducing its computational burden.

[0037] This invention actively models and suppresses noise at its physical source, fundamentally avoiding signal damage caused by overlap between noise and useful signal frequency bands in traditional filtering methods, thus significantly improving noise reduction accuracy and signal-to-noise ratio. Simultaneously, a two-stage filtering architecture is formed by combining pre-comb filtering with adaptive fine filtering, transforming the computationally intensive adaptive filtering object from the original signal to a pre-purified signal, greatly reducing the overhead on embedded hardware resources and achieving a balance between excellent noise reduction performance and low computational cost. Furthermore, the entire process can respond to changes in vehicle status in real time, dynamically adjusting parameters and possessing continuous adaptive capabilities, providing a reliable, high-quality signal source for core in-vehicle physiological monitoring functions.

[0038] Based on the above technical solutions, the embodiments of the present invention can be further improved as follows.

[0039] In one possible embodiment, in step S1, the raw physiological signal is acquired by a capacitive sensor integrated into the steering wheel or seat, and the vehicle status data is acquired in real time through the vehicle bus system and inertial measurement unit.

[0040] For example, when the driver holds the steering wheel with a built-in capacitive sensor, the sensor simultaneously collects a mixed electrical signal containing electrocardiogram waveforms and vehicle vibration noise. At the same time, the system reads engine speed (e.g., 2500 rpm) and vehicle speed (e.g., 80 km / h) in real time via the vehicle's CAN bus, and acquires vibration data such as vehicle acceleration through the onboard IMU. This embodiment, through hardware-level collaborative design, strictly aligns physiological signals with the vehicle state parameters that generate noise based on a timing synchronization mechanism, ensuring the time alignment accuracy of multi-source data. This provides a real and synchronized data foundation for accurately generating noise reference signals in subsequent steps, fundamentally ensuring the accuracy of the noise reduction model.

[0041] In one possible embodiment, in step S2, the main characteristic frequencies of the current vehicle vibration are calculated based on the vehicle state data, wherein the main characteristic frequencies include at least the fundamental frequency of the engine vibration and the characteristic frequency of the road surface vibration, and are obtained as follows: (1) Based on the engine speed, the fundamental frequency of engine vibration is calculated using the following formula. :

[0042] in: RPM is the engine speed, obtained from the CAN bus; Number of engine cylinders (system preset); This is the ignition coefficient (0.5 for a four-stroke engine). For gearbox ratio (obtained in real time); This is the correction factor for the current gear.

[0043] (2) Based on the vehicle speed, calculate the road vibration characteristic frequency using the following formula. :

[0044] in: Vehicle speed (km / h); The diameter of the tire; This refers to the pavement correction term based on IMU spectral analysis.

[0045] In a specific implementation scenario, when the vehicle is traveling at 80 km / h and the engine speed is 2500 rpm, the system first calculates the dominant fundamental frequency of the current engine vibration based on the engine speed data (2500 rpm), preset four-cylinder engine parameters, and transmission ratio, using the engine vibration fundamental frequency calculation formula. (e.g., approximately 41.7 Hz).

[0046] Simultaneously, based on the current vehicle speed (80km / h) and tire diameter, the main vibration frequencies caused by road excitation are calculated using the road vibration characteristic frequency calculation formula. (e.g., approximately 12Hz).

[0047] The system uses these two characteristic frequencies (41.7Hz and 12Hz) derived from the powertrain and road surface excitation, respectively, as the basis for the current main vibration and noise model, providing an accurate frequency core for the subsequent generation of a complete noise reference signal containing its harmonic components.

[0048] In one possible embodiment, step S2, generating the corresponding noise reference signal, includes: The engine vibration fundamental frequency ( ) and its harmonics, and the characteristic frequency of the road surface vibration ( The noise reference signal is generated by synthesizing its harmonics. r(t) ; Wherein, the noise reference signal r(t) Represented as the superposition of multiple sine wave components:

[0049] Where n is the harmonic order of the harmonic component corresponding to the fundamental frequency of engine vibration. This represents the amplitude of the harmonic component corresponding to the fundamental frequency of engine vibration. The phase of the harmonic component corresponding to the fundamental frequency of engine vibration; m represents the harmonic order of the harmonic component corresponding to the characteristic frequency of road vibration. The amplitude of the harmonic component corresponding to the characteristic frequency of road vibration. This represents the phase of the harmonic component corresponding to the characteristic frequency of road vibration.

[0050] It is understood that this embodiment uses noise harmonic synthesis technology to transform discrete frequency parameters into a high-fidelity noise time-domain model. This embodiment is based on the calculated engine vibration fundamental frequency. and road surface vibration characteristic frequency It not only utilizes its fundamental component, but more importantly, it introduces and synthesizes its various harmonics through formulas. , This allows for the construction of a complete digital noise reference signal that includes the main spectral characteristics of the noise. .

[0051] This embodiment greatly improves the accuracy and efficiency of subsequent adaptive filtering: Since vehicle vibration noise is not a simple sine wave in reality, but a composite signal containing rich harmonic components, the noise reference signal r(t) generated in this embodiment is closer to the real physical noise in terms of waveform and spectrum. This allows the subsequent adaptive filter to efficiently remove all frequency components highly correlated with the noise source from the physiological signal, achieving a leap from coarse frequency matching to fine waveform cancellation, and effectively improving the noise reduction depth and signal fidelity under complex working conditions.

[0052] In one possible embodiment, step S3 includes: First, based on the main characteristic frequencies and the system sampling rate, the notch depth control parameter N of the comb filter is calculated; Secondly, the coefficients of the programmable comb filter are configured according to the parameter N; Specifically, the transfer function of the comb filter is expressed as:

[0053] in: N is the notch depth control parameter. , The system sampling rate, For the center frequency, Obtained through the aforementioned main characteristic frequencies; This indicates a signal delay of N sampling points; The pole radius is used to control the notch depth and bandwidth.

[0054] The notch depth control parameter N and the pole radius Adaptive adjustments are made based on the vehicle's real-time status.

[0055] Then, the original physiological signal is filtered using a configured comb filter to obtain the preliminary noise-reduced signal.

[0056] For example, when the system calculates the current engine vibration fundamental frequency based on the vehicle status... At 41.7Hz, firstly, based on the system sampling rate... (e.g., 1000Hz) Execute step S3 to calculate the key parameters of the comb filter N = round(1000 / 41.7) ≈ 24; then, according to the defined transfer function... Configure a comb filter with a center frequency of 41.7Hz, generating periodic notch filters at its harmonics (83.4Hz, 125.1Hz, etc.); where the pole radius... Based on the real-time signal-to-noise ratio being adaptively set to 0.97, the notch filter is controlled to have an appropriate depth and bandwidth. Ultimately, after the original physiological signal passes through this precisely tuned filter, the vibration noise and its harmonics strongly correlated with 41.7Hz are significantly weakened, and the output signal is initially purified, creating favorable conditions for subsequent adaptive filtering.

[0057] In one possible embodiment, step S3, which involves adaptively adjusting the notch depth control parameter N based on the real-time vehicle status, includes: The current engine vibration fundamental frequency is dynamically calculated based on real-time vehicle status data. And based on the current engine vibration fundamental frequency Update the notch depth control parameter N to make the notch center frequency of the comb filter... The formula for real-time adjustment based on changes in vehicle operating conditions is as follows:

[0058] in, The notch depth control parameters are updated adaptively.

[0059] Understandably, this embodiment solves the noise frequency drift problem caused by changes in vehicle operating conditions by designing the notch depth control parameter N of the comb filter as an adaptive variable dynamically updated based on the real-time engine frequency. This enables the noise reduction system to have real-time tracking capabilities. When the engine speed changes due to vehicle acceleration, deceleration, or gear shifting, this embodiment can immediately calculate the new noise fundamental frequency and synchronously adjust the comb filter parameters. This ensures that the notch center frequency is always accurately locked onto the most significant vibration noise, avoiding the drawback of traditional fixed-parameter filters failing rapidly under changing operating conditions. Thus, it maintains stable and high-precision noise reduction performance continuously in dynamic driving environments.

[0060] In one possible embodiment, in step S3, the radius of the pole is adaptively adjusted according to the real-time state of the vehicle. The steps include: Signal-to-noise ratio based on real-time estimation The pole radius is updated to dynamically optimize the notch depth and bandwidth of the comb filter based on signal quality. The formula for updating the pole radius is:

[0061] in, The updated pole radius, This is the preset maximum signal-to-noise ratio.

[0062] This embodiment establishes the pole radius using the above formula. With real-time signal-to-noise ratio The adaptive correlation enables intelligent dynamic optimization of the comb filter's notch characteristics, thereby achieving an optimal balance between noise reduction strength and signal fidelity. When the signal quality is poor (corresponding to a low signal-to-noise ratio), the system automatically reduces... The value is increased to widen the notch bandwidth, ensuring effective capture and filtering of strong noise that may be accompanied by frequency drift, preventing missed detections; when the signal quality is good (corresponding to a high signal-to-noise ratio), the system increases... The notch bandwidth is narrowed to minimize damage to useful physiological waveforms (such as the ECG R wave). The adaptive mechanism in this embodiment enables the filter to flexibly adjust its strategy according to real-time signal conditions, significantly improving the robustness and reliability of the system in various complex vehicle environments.

[0063] In one possible embodiment, step S4 includes: The initial noise reduction signal is used as the main input. The noise reference signal is used as the reference input. They are input together into the adaptive filter; The adaptive filter is based on its current weight vector. For reference input Perform weighted calculations to obtain the filter output. ; Calculate the main input With the filter output Error between The error This refers to the physiological signal after final noise reduction; Based on the error and the reference input The filter weight vector is updated using an adaptive algorithm. It is used for subsequent signal filtering.

[0064] The Least Mean Square (LMS) adaptive filter will be used as an example for illustration.

[0065] For example, after initial noise reduction via comb filtering, the preliminary ECG signal containing residual vibration noise is used as the main input. The noise reference signal, which includes fundamental and harmonic frequencies, is generated by engine speed (2500 rpm) and vehicle speed (80 km / h). As a reference input The common input is the LMS adaptive filter; the filter passes through the current weight vector. The noise reference signal is weighted and calculated to output a predicted noise that is most similar to the actual noise waveform mixed in the ECG signal. ; will the main input With prediction noise Error signal obtained by subtraction This is a pure electrocardiogram signal. At this point, the system automatically updates the filter weights by monitoring the error magnitude. This prepares the signal for processing the next sampling point, enabling continuous and accurate tracking and elimination of noise.

[0066] Understandably, step S4 constitutes a complete adaptive noise cancellation closed-loop system with continuous learning capabilities. It not only achieves instantaneous and accurate noise cancellation but also, through real-time weight updates, enables the system to track the time-varying characteristics of noise. The dynamic feedback adjustment mechanism of this embodiment ensures that, when faced with constantly changing vibration noise due to vehicle acceleration, deceleration, or road condition changes, the adaptive filter can continuously optimize its internal model, always maintaining optimal noise reduction performance, thus improving the robustness, stability, and long-term effectiveness of the entire noise reduction scheme in non-stationary vehicle environments.

[0067] As an optional implementation, step S4, the step of adaptively adjusting the adaptive filter, further includes: In the adaptive algorithm, the step size factor A variable step size strategy is adopted, the value of which is based on the reference input signal. The instantaneous energy is adaptively adjusted, and the calculation formula is:

[0068] Where β is the convergence rate factor and α is the stabilization constant; Furthermore, the order L of the adaptive filter is based on the current noise fundamental frequency. and system sampling rate After optimization, the optimized order is:

[0069] Wherein, ΔL is the order fine-tuning amount preset based on experience.

[0070] The adaptive algorithm in this embodiment clarifies that the adaptive filter does not use traditional fixed parameters, but has dual adaptive capabilities: firstly, by varying the step size... Achieving the optimal balance between convergence speed and steady-state accuracy; secondly, by optimizing the order. This approach matches the filter complexity to the current noise characteristics. This embodiment improves the algorithm's convergence, stability, and computational efficiency in non-stationary on-board environments.

[0071] In one possible embodiment, the present invention also employs a circular buffer mechanism for memory optimization design, which is expressed as follows:

[0072] in, The length of historical data required for the adaptive filter (i.e., its order). Let be the delay parameter of the comb filter, which is represented in the transfer function of the comb filter as follows: , indicating that a delay of N sampling points is required; This indicates that the maximum value of the two values ​​is taken. This ensures that the allocated memory buffer is large enough to simultaneously meet the historical data length requirement of the filter with the higher demand.

[0073] × 2 (double), this is a key safety redundancy design. Allocating twice the maximum memory requirement creates a "safe zone".

[0074] This embodiment ensures data processing continuity through a double-buffered design. A specific memory management strategy minimizes memory overhead while meeting the stringent requirements of signal processing algorithms for data continuity and real-time performance. This embodiment resolves the contradiction in real-time streaming signal processing where new data continuously arrives while older data needs to be processed and overwritten promptly.

[0075] To verify the noise reduction effect of the technical solution of the present invention, a test was conducted on a simulated vehicle vibration table.

[0076] Test conditions: Under simulated vibration conditions with a vehicle speed of 80 km / h and an engine speed of 2500 rpm, electrocardiogram signals of a simulated driver were collected using capacitive electrodes.

[0077] Evaluation metrics: Signal-to-noise ratio (SNR) and heart rate detection accuracy.

[0078] The evaluation results are shown in Table 1: Table 1 Evaluation Results

[0079] As shown in Table 1, under simulated vehicle vibration conditions with a vehicle speed of 80 km / h and an engine speed of 2500 rpm, the signal-to-noise ratio of the output signal significantly improved from 5.2 dB in the traditional method (50 Hz notch filter + fixed bandwidth bandpass filter) to 15.8 dB after applying the directional noise reduction scheme of this invention. At the same time, the accuracy of heart rate detection increased significantly from 78.5% to 98.7%. This quantitative comparison result directly verifies the effectiveness of the vehicle state data-driven and two-stage collaborative filtering scheme in this invention, proving that it can accurately suppress specific vibration noise that is strongly correlated with vehicle operating conditions, thus providing solid experimental evidence that the accuracy and reliability of core physiological parameter monitoring have reached the practical level.

[0080] More detailed performance test data can be found in Table 2: Table 2 Performance under different operating conditions

[0081] As shown in Table 2, under various conditions ranging from idling to high-speed cruising and even rough, bumpy road surfaces, the signal-to-noise ratio (SNR) and R-wave detection rate of this invention are comprehensively and significantly superior to traditional methods. Especially in the most complex noise conditions of "urban roads" and "bumpy road surfaces," the SNR improvement is the greatest (reaching +10.6 dB and +8.7 dB respectively), and the R-wave detection rate also improves from a severely inaccurate level (78.5% and 65.3%) to a usable or even excellent level (98.7% and 94.2%). This strongly demonstrates that the vehicle state-based adaptive noise reduction mechanism of this invention can effectively cope with various dynamically changing driving environments, ensuring the reliability and stability of the system in practical applications.

[0082] Test conclusion: The directional noise reduction scheme of the present invention can significantly improve the signal-to-noise ratio of the output signal by accurately locating and suppressing vibration noise using vehicle status information, thereby enabling the accuracy of heart rate detection to reach a practical level under severe vehicle vibration environment.

[0083] Figure 2 This is a structural diagram of a directional noise reduction system for vehicle-mounted capacitive physiological signal monitoring provided by an embodiment of the present invention, as shown below. Figure 2 As shown, a directional noise reduction system for vehicle-mounted capacitive physiological signal monitoring includes a signal acquisition module, a noise reference signal generation module, a programmable comb filter, and an adaptive filtering module, wherein: The signal acquisition module is used to acquire the driver's raw physiological signals and vehicle status data in real time, wherein the vehicle status data includes at least engine speed and vehicle speed. A noise reference signal generation module, connected to the signal acquisition module, is used to calculate the main characteristic frequencies of the current vehicle vibration based on the vehicle state data and generate the corresponding noise reference signal. A programmable comb filter, connected to the signal acquisition module and the noise reference signal generation module, is used to filter the original physiological signal with the characteristic frequency as the center to obtain a preliminary noise-reduced signal; An adaptive filtering module is connected to the programmable comb filter and the noise reference signal generation module, respectively, and is used to filter the preliminary noise reduction signal based on the noise reference signal to output the final noise-reduced physiological signal.

[0084] It is understood that the directional noise reduction system for vehicle-mounted capacitive physiological signal monitoring provided by the present invention corresponds to the directional noise reduction method for vehicle-mounted capacitive physiological signal monitoring provided in the foregoing embodiments. The relevant technical features of the directional noise reduction system for vehicle-mounted capacitive physiological signal monitoring can be referred to the relevant technical features of the directional noise reduction method for vehicle-mounted capacitive physiological signal monitoring, and will not be repeated here.

[0085] Please see Figure 3 , Figure 3 A schematic diagram illustrating an embodiment of the electronic device provided in this invention. For example... Figure 3 As shown, this embodiment of the invention provides an electronic device 300, including a memory 310, a processor 320, and a computer program 311 stored in the memory 310 and executable on the processor 320. When the processor 320 executes the computer program 311, it performs the following steps: S1, real-time acquisition of the driver's raw physiological signals and vehicle status data, wherein the vehicle status data includes at least engine speed and vehicle speed; S2, Based on the vehicle status data, calculate the main characteristic frequencies of the current vehicle vibration and generate the corresponding noise reference signal; S3, using the main characteristic frequency as the center frequency, perform programmable comb filtering on the original physiological signal to obtain a preliminary noise-reduced signal; S4, based on the noise reference signal, the preliminary noise reduction signal is adaptively filtered to obtain the final noise-reduced physiological signal.

[0086] Please see Figure 4 , Figure 4 This is a schematic diagram illustrating an embodiment of a computer-readable storage medium provided by the present invention. (See diagram below.) Figure 4 As shown, this embodiment provides a computer-readable storage medium 400 on which a computer program 311 is stored. When the computer program 311 is executed by a processor, it performs the following steps: S1, real-time acquisition of the driver's raw physiological signals and vehicle status data, wherein the vehicle status data includes at least engine speed and vehicle speed; S2, Based on the vehicle status data, calculate the main characteristic frequencies of the current vehicle vibration and generate the corresponding noise reference signal; S3, using the main characteristic frequency as the center frequency, perform programmable comb filtering on the original physiological signal to obtain a preliminary noise-reduced signal; S4, based on the noise reference signal, the preliminary noise reduction signal is adaptively filtered to obtain the final noise-reduced physiological signal.

[0087] This invention provides a directional noise reduction method, system, and storage medium for vehicle-mounted capacitive physiological signal monitoring, constructing a closed-loop directional noise cancellation system with "feedforward-feedback coordination." This invention first utilizes vehicle CAN bus and IMU data as feedforward signals, shifting the noise reduction focus from traditional back-end signal processing to the physical source of the noise. Specifically, the system uses a built-in physical model (such as the engine ignition frequency formula) to convert state parameters such as engine speed and vehicle speed into a precise digital model of the noise in real time (i.e., a noise reference signal). Subsequently, the system adopts a two-stage filtering architecture of "dedicated preprocessing + general fine processing": First, a programmable comb filter with dynamic configuration based on the frequency parameters of the noise model is used to perform fast and efficient directional filtering on the original physiological signal with its unique periodic frequency notch characteristics, specifically eliminating discrete spectral noise that is strongly correlated with vehicle vibration; then, the signal after preliminary noise reduction and the noise reference signal are fed into an improved adaptive filter (using the variable step size LMS algorithm). This filter dynamically adjusts the weights in a feedback manner by continuously comparing the output error with the reference input, thereby achieving fine tracking and cancellation of residual noise and time-varying noise components.

[0088] This invention achieves three significant technical effects through a two-tier collaborative architecture of vehicle state data-driven noise source modeling and comb filtering / adaptive filtering: In terms of noise reduction performance, it can accurately suppress vehicle vibration noise that overlaps with physiological signal frequency bands, improving the signal-to-noise ratio by more than 10dB, and increasing the heart rate detection accuracy from 78.5% in traditional methods to a practical level of 98.7%; In terms of computational efficiency, by distributing the main computational load through preprocessing with comb filtering, the system processing latency is optimized from 45ms to 18ms, and the CPU utilization rate is reduced to 32%, meeting the real-time requirements of the vehicle-mounted embedded platform; In terms of system adaptability, thanks to the real-time adaptive parameter adjustment mechanism, it ensures stable noise reduction robustness under dynamic conditions such as acceleration, deceleration, and bumpy roads, providing a highly reliable signal foundation for the core vehicle health monitoring system.

[0089] It should be noted that the descriptions of each embodiment in the above embodiments have different focuses. For parts that are not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.

[0090] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0091] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded computer, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0092] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0093] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0094] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including both the preferred embodiments and all changes and modifications falling within the scope of the invention.

[0095] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A directional noise reduction method for vehicle-mounted capacitive physiological signal monitoring, characterized in that, include: S1, real-time acquisition of the driver's raw physiological signals and vehicle status data, wherein the vehicle status data includes at least engine speed and vehicle speed; S2, Based on the vehicle status data, calculate the main characteristic frequencies of the current vehicle vibration and generate the corresponding noise reference signal; S3, using the main characteristic frequency as the center frequency, perform programmable comb filtering on the original physiological signal to obtain a preliminary noise-reduced signal; S4, based on the noise reference signal, the preliminary noise reduction signal is adaptively filtered to obtain the final noise-reduced physiological signal.

2. The directional noise reduction method for vehicle-mounted capacitive physiological signal monitoring according to claim 1, characterized in that, In step S1, the raw physiological signals are acquired by capacitive sensors integrated into the steering wheel or seat, and the vehicle status data is acquired in real time through the vehicle bus system and inertial measurement unit.

3. The directional noise reduction method for vehicle-mounted capacitive physiological signal monitoring according to claim 1, characterized in that, In step S2, calculating the main characteristic frequencies of the current vehicle vibration based on the vehicle state data includes: Calculate the engine vibration fundamental frequency based on the engine speed; Based on the vehicle speed, calculate the characteristic frequency of road vibration; The main characteristic frequencies include at least the fundamental frequency of engine vibration and the characteristic frequency of road surface vibration.

4. A directional noise reduction method for vehicle-mounted capacitive physiological signal monitoring according to claim 3, characterized in that, In step S2, generating the corresponding noise reference signal includes: The engine vibration fundamental frequency ( ) and its harmonics, and the characteristic frequency of the road surface vibration ( The noise reference signal is generated by synthesizing its harmonics. ; Wherein, the noise reference signal Represented as the superposition of multiple sine wave components: Where n and m are the harmonic orders, and For the amplitude of the corresponding harmonic components, and This represents the phase of the corresponding harmonic component.

5. A directional noise reduction method for vehicle-mounted capacitive physiological signal monitoring according to claim 1, characterized in that, Step S3 includes: Based on the main characteristic frequencies and the system sampling rate, the notch depth control parameter N of the comb filter is calculated; Configure the coefficients of the programmable comb filter according to the parameter N; The original physiological signal is filtered using a pre-configured comb filter to obtain the preliminary noise-reduced signal.

6. A directional noise reduction method for vehicle-mounted capacitive physiological signal monitoring according to claim 5, characterized in that, The transfer function of the comb filter is expressed as: in: N is the notch depth control parameter. , The system sampling rate, For the center frequency, Obtained through the aforementioned main characteristic frequencies; This indicates a signal delay of N sampling points; The pole radius is used to control the notch depth and bandwidth; The notch depth control parameter N and the pole radius Adaptive adjustments are made based on the vehicle's real-time status.

7. A directional noise reduction method for vehicle-mounted capacitive physiological signal monitoring according to claim 6, characterized in that, The step of adaptively adjusting the notch depth control parameter N based on the real-time vehicle status includes: The current engine vibration fundamental frequency is dynamically calculated based on real-time vehicle status data. And based on the current engine vibration fundamental frequency Update the notch depth control parameter N to make the notch center frequency of the comb filter... The formula for real-time adjustment based on changes in vehicle operating conditions is as follows: in, The notch depth control parameters are updated adaptively.

8. A directional noise reduction method for vehicle-mounted capacitive physiological signal monitoring according to claim 6, characterized in that, The radius of the pole is adaptively adjusted according to the real-time status of the vehicle. The steps include: Signal-to-noise ratio based on real-time estimation The pole radius is updated to dynamically optimize the notch depth and bandwidth of the comb filter based on signal quality. The formula for updating the pole radius is: in, The updated pole radius, This is the preset maximum signal-to-noise ratio.

9. A directional noise reduction method for vehicle-mounted capacitive physiological signal monitoring according to claim 1, characterized in that, Step S4 includes: The initial noise reduction signal is used as the main input. The noise reference signal is used as the reference input. They are input together into the adaptive filter; The adaptive filter is based on its current weight vector. For reference input Perform weighted calculations to obtain the filter output. ; Calculate the main input With the filter output Error between The error This refers to the physiological signal after final noise reduction; Based on the error and the reference input The filter weight vector is updated using an adaptive algorithm. It is used for subsequent signal filtering.

10. A directional noise reduction method for vehicle-mounted capacitive physiological signal monitoring according to claim 9, characterized in that, Step S4, the step of adaptively adjusting the adaptive filter, further includes: The adaptive algorithm employs a variable step size strategy and a step size factor. The value depends on the reference input signal The instantaneous energy is adaptively adjusted, and the calculation formula is: Where β is the convergence rate factor and α is the stabilization constant; Furthermore, the order L of the adaptive filter is based on the current noise fundamental frequency. and system sampling rate After optimization, the optimized order is expressed as: in, This is a fine-tuning amount based on a preset order from experience.

Images