Automobile driver state monitoring method based on physiological sign extraction

Abstract

The invention relates to the technical field of automobile driver state monitoring, and discloses an automobile driver state monitoring method based on physiological sign extraction. The method comprises the following steps: continuously capturing heart rate, breath and posture movement signal streams of a driver; respectively carrying out time-frequency analysis on the heart rate signal to obtain a heart rate variability quantitative spectrogram, decomposing and correcting the respiratory signal to extract a respiratory depth change envelope, and carrying out trajectory reconstruction and frequency division on the attitude signal to obtain a body movement characteristic frequency set; constructing a multi-dimensional physiological feature matrix based on the features, performing layer-by-layer matching and difference calculation on the multi-dimensional physiological feature matrix and a preset state template, and generating a state deviation degree sequence and a physiological risk mark; and selecting an initial intervention instruction set according to the mark, fusing the deviation degree sequence to dynamically calibrate parameters of the initial intervention instruction set, and generating a final intervention instruction set. According to the method, modeled accurate recognition and self-adaptive dynamic intervention of the driver state are realized, and the monitoring accuracy and the intervention effectiveness are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of automobile driver state monitoring, in particular to an automobile driver state monitoring method based on physiological sign extraction. BACKGROUND

[0002] At present, driver state monitoring mostly relies on threshold judgment of single or limited signal sources. For example, fatigue is determined by image recognition of facial features, or vehicle behavior parameters such as steering wheel angle and lane deviation are monitored. This kind of method usually independently collects and analyzes physiological signals such as heart rate and respiration, and only compares their values with preset fixed safety thresholds. The corresponding intervention mechanism is mostly static response, that is, when the alarm condition is triggered, the preset uniform warning action is executed.

[0003] This judgment mode based on isolated signals and fixed thresholds cannot accurately reflect the complex state of the driver composed of the coordinated changes of multiple physiological signs. It is easily affected by individual differences and instantaneous interference, resulting in frequent false positives and false negatives. At the same time, the rigid intervention strategy cannot respond differently according to the actual degree and trend of state deterioration, which may fail due to insufficient warning or cause interference due to excessive warning.

[0004] A technical solution is needed that can comprehensively utilize multi-dimensional physiological signs, conduct overall patternized evaluation of the driver's state, and dynamically adjust the intervention strategy according to the continuous trend of state changes, in order to overcome the shortcomings of existing methods in state recognition accuracy and intervention measure adaptability. SUMMARY

[0005] The purpose of the present application is to provide an automobile driver state monitoring method based on physiological sign extraction to solve the problems raised in the background art.

[0006] To achieve the above purpose, the present application provides an automobile driver state monitoring method based on physiological sign extraction, which comprises: continuously capturing a set of original physiological information of a target driver, the set of original physiological information comprising a heart rate signal stream, a respiration signal stream, and a posture motion signal stream; performing time domain pattern separation and frequency domain spectrum analysis on the heart rate signal stream to obtain a reference waveform of heart rate rhythm and a quantized spectrum graph of heart rate variability; performing non-stationary signal decomposition and baseline drift elimination on the respiration signal stream to extract a respiration fundamental frequency waveform and a change envelope line of respiration depth; performing three-dimensional space trajectory reconstruction and frequency component division on the posture motion signal stream to obtain a distribution spectrum of body movement energy on each axis in space and a characteristic frequency set of periodic motion; constructing a multi-dimensional physiological feature matrix based on the quantized spectrum graph, the change envelope line, and the characteristic frequency set; matching and difference calculation with preset driver state templates, to generate a state deviation sequence and a plurality of physiological risk markers; According to the physiological risk markers, select the corresponding initial intervention instruction set from the preset intervention strategy library; Fusion of the state deviation sequence to dynamically calibrate the parameters of the initial intervention instruction set to generate a final intervention instruction set.

[0007] Preferably, the time domain shape separation and frequency domain spectrum analysis are performed on the heart rate signal stream to obtain the reference waveform of heart rate rhythm and the quantization spectrum of heart rate variability, including: A multi-scale sliding window is applied to the heart rate signal stream to obtain a series of heart rate signal segments; Local extreme point detection is performed on each of the heart rate signal segments to determine the time coordinates of the start point, peak point and end point of each heart cycle; According to the time coordinates, the waveforms of all heart cycles are aligned and superimposed averaged to obtain the reference waveform of the heart rate rhythm; Subtract the reference waveform from each of the heart rate signal segments to obtain a heart rate residual signal; Power spectral density estimation is performed on the heart rate residual signal to calculate the energy proportion in the ultra-low frequency, low frequency and high frequency bands, forming the quantization spectrum of the heart rate variability.

[0008] Preferably, the non-stationary signal decomposition and baseline drift elimination are performed on the respiratory signal stream to extract the respiratory fundamental frequency waveform and the change envelope of respiratory depth, including: Adaptive mode decomposition is applied to the respiratory signal stream to decompose it into a plurality of intrinsic mode function components and a residual trend component; Identify the target component whose energy is concentrated near the respiratory frequency among the intrinsic mode function components, and reconstruct the target component into a pure respiratory signal; Subtract the residual trend component from the respiratory signal stream to obtain a respiratory corrected signal with slow trend removed; Calculate the instantaneous frequency of the pure respiratory signal to obtain the waveform of the respiratory fundamental frequency changing with time; Absolute value processing and low-pass filtering are performed on the respiratory corrected signal to obtain the change envelope of the respiratory depth representing the slow change of the respiratory amplitude.

[0009] Preferably, the three-dimensional space trajectory reconstruction and frequency component division are performed on the posture motion signal stream to obtain the distribution spectrum of body movement energy on each axis in space and the characteristic frequency set of periodic motion, including: separating three-dimensional linear acceleration data and three-dimensional angular velocity data from the attitude motion signal stream; secondly integrating the three-dimensional linear acceleration data to obtain a displacement trajectory of a body part of the driver in a three-dimensional space by attitude solution in combination with the three-dimensional angular velocity data; respectively performing Fourier transform on the projection components of the displacement trajectory on each axis of a space orthogonal coordinate system to obtain a frequency spectrum in each axis direction, and a distribution spectrum of the body motion energy on each axis of the space being constituted by the frequency spectrum; detecting a significant peak value in the frequency spectrum in each axis direction, comparing a frequency corresponding to the peak value with a preset typical body motion frequency range, screening out a frequency point belonging to periodic motion to constitute the characteristic frequency set.

[0010] Preferably, the multi-dimensional physiological feature matrix is constructed based on the quantized spectrum graph, the change envelope line and the characteristic frequency set, comprising: extracting an ultra-low frequency energy proportion, a low frequency to high frequency energy ratio and a total power value from the quantized spectrum graph as a first group of feature vectors; extracting a mean value, a variance, a slope and a zero-crossing rate from the change envelope line as a second group of feature vectors; extracting a main frequency value, a frequency concentration degree and a frequency stability coefficient from the characteristic frequency set as a third group of feature vectors; splicing the first group of feature vectors, the second group of feature vectors and the third group of feature vectors in a predetermined order to form a row vector; repeating the above steps on a plurality of continuous time frames, stacking the obtained plurality of row vectors in time sequence to form the multi-dimensional physiological feature matrix, wherein each row of the matrix corresponds to the multi-dimensional features of a time frame.

[0011] Preferably, the multi-dimensional physiological feature matrix is matched with a preset driver state template layer by layer and a difference degree is calculated to generate a state deviation degree sequence and a plurality of physiological risk markers, comprising: obtaining the preset driver state template, the preset driver state template including a normal state feature template, a fatigue state feature template and a disease risk state feature template; calculating the Euclidean distance between each row feature vector of the multi-dimensional physiological feature matrix and the normal state feature template to form a basic deviation degree sequence; calculating the similarity between each row feature vector of the multi-dimensional physiological feature matrix and the fatigue state feature template, and converting the similarity into a fatigue deviation degree; calculating a matching score between each row feature vector of the multi-dimensional physiological feature matrix and a specific risk pattern in the disease risk state feature template; a distraction risk flag is generated when the value of the base deviation sequence continuously exceeds a warning threshold; a fatigue accumulation risk flag is generated when the fatigue deviation continuously exceeds a fatigue threshold; a sudden health risk flag is generated when the matching score of a specific risk pattern exceeds a risk threshold.

[0012] Preferably, the initial intervention instruction set is selected from a preset intervention strategy library according to the physiological risk flag, including: in response to the distraction risk flag, a distraction awakening strategy is called from the intervention strategy library, the distraction awakening strategy including initial parameters of acoustic stimulation type, visual prompt content, and trigger duration; in response to the fatigue accumulation risk flag, an anti-fatigue strategy is called from the intervention strategy library, the anti-fatigue strategy including initial parameters of recommended rest interval, seat massage mode, and air conditioning adjustment target; in response to the sudden health risk flag, an emergency response strategy is called from the intervention strategy library, the emergency response strategy including initial parameters of vehicle control recommendation, contact priority, and information broadcast template; the initial parameters of the selected strategy are summarized to form the initial intervention instruction set.

[0013] Preferably, the parameters of the initial intervention instruction set are dynamically calibrated by fusing the state deviation sequence to generate a final intervention instruction set, including: the latest value and change trend of the state deviation sequence are read; the intensity parameter of acoustic stimulation or the frequency parameter of visual prompt in the initial intervention instruction set is linearly adjusted according to the size of the latest value; the duration parameter of recommended rest interval in the initial intervention instruction set is nonlinearly adjusted according to the slope of the change trend; the intensity curve of seat massage mode and the specific temperature value of air conditioning adjustment target are weightedly corrected in combination with the current time, historical driving time, and environmental light intensity; all the parameters adjusted in intensity, frequency, duration, and specific target value are integrated to form the final intervention instruction set that can be directly parsed by a vehicle execution system.

[0014] Preferably, after the final intervention instruction set is generated, a model adaptive updating step is further included: the feedback physiological information set of the target driver after implementation of the intervention measure is received in real time; the heart rate variability quantization spectrum, the respiration depth change envelope, and the body motion characteristic frequency set of the subsequent time period are extracted from the feedback physiological information set to form a feedback feature matrix; calculating a feature drift between the feedback feature matrix and a multi-dimensional physiological feature matrix on which the final intervention instruction set is generated; if the feature drift is greater than an adaptive update threshold, then using the feedback feature matrix to incrementally update a corresponding state feature template in the preset driver state template.

[0015] Preferably, the continuously captured original physiological information set of the target driver includes: embedding a distributed inertial sensing unit in the backrest and cushion of the vehicle seat to generate original motion data of the posture motion signal stream by measuring the three-dimensional acceleration and angular velocity of the driver's torso and hips; embedding a distributed inertial sensing unit in the backrest and cushion of the vehicle seat to generate original motion data of the posture motion signal stream by measuring the three-dimensional acceleration and angular velocity of the driver's torso and hips; using a high-precision analog-to-digital converter to synchronously sample and digitize the original electrical signal and the original motion data, ensuring that the heart rate signal stream, the respiratory signal stream, and the posture motion signal stream have a unified and synchronized time reference.

[0016] Compared with the prior art, the present application has the following advantages: By fusing multi-dimensional heterogeneous physiological features such as heart rate variability spectrogram, respiratory depth envelope, and body motion feature frequency into a feature matrix, and performing layer-by-layer matching and difference calculation with the preset driver state template, pattern recognition of the driver's state is realized. This method changes from isolated threshold judgment to multi-dimensional and overall pattern similarity evaluation, which can more accurately capture complex states represented by the coordinated changes of multiple signs, reduce the misjudgment rate caused by single physiological signal fluctuation or environmental interference, and improve the robustness and accuracy of state recognition. The generated continuous state deviation sequence can quantitatively reflect the trend and degree of state deviation from the normal pattern, providing continuous parameter basis for dynamic assessment of the state.

[0017] According to the state deviation sequence, the initial intervention instruction parameters selected from the strategy library are dynamically calibrated, so that the final executed intervention instruction has adaptive characteristics. This makes the intensity, frequency or other parameters of the intervention be able to be nonlinearly adjusted according to the real-time state deviation degree and its change trend of the driver, rather than executing fixed instructions. The dynamic calibration mechanism realizes the individualization and gradientization of the intervention strategy, which can take soft prompt for mild state deviation and automatically enhance the intervention intensity for sharp or severe deviation, thereby ensuring the effectiveness of the warning while reducing unnecessary interference to the driver, improving the intelligence and acceptance of human-computer interaction. BRIEF DESCRIPTION OF DRAWINGS

[0018] Figure 1The working principle diagram of the automobile driver state monitoring method based on physiological sign extraction according to the application; Figure 2 The flow chart for heart rate signal processing and quantitative spectrum generation; Figure 3 The flow chart for multi-dimensional physiological feature matrix construction; Figure 4 The nonlinear relationship diagram of the recommended rest interval and the state deviation degree slope; Figure 5 The diagram of the feature drift amount change over time and the adaptive update threshold association. DETAILED DESCRIPTION

[0019] The technical solutions in the embodiments of the application will be described clearly and completely below with reference to the drawings in the embodiments of the application. Obviously, the described embodiments are only part of the embodiments of the application, rather than all the embodiments of the application. Based on the embodiments in the application, all other embodiments obtained by a person of ordinary skill in the art without creative labor fall within the protection scope of the application.

[0020] Please refer to Figure 1 The application provides an automobile driver state monitoring method based on physiological sign extraction. The method comprises the following steps: continuously capturing a raw physiological information set of a target driver, the set comprising a heart rate signal stream, a respiration signal stream, and a posture motion signal stream; performing time domain shape separation and frequency domain spectrum analysis on the heart rate signal stream, so as to obtain a reference waveform of heart rate rhythm and a quantitative spectrum graph of heart rate variability; performing non-stationary signal decomposition and baseline drift elimination on the respiration signal stream, so as to extract a respiration fundamental frequency waveform and a change envelope line of respiration depth; performing three-dimensional space trajectory reconstruction and frequency component division on the posture motion signal stream, so as to obtain a distribution spectrum of body movement energy on each axis in space and a characteristic frequency set of periodic motion; constructing a multi-dimensional physiological feature matrix based on the obtained quantitative spectrum graph, change envelope line, and characteristic frequency set; performing layer-by-layer matching and difference degree calculation on the constructed multi-dimensional physiological feature matrix and a preset driver state template, so as to generate a state deviation degree sequence and a plurality of physiological risk markers; selecting a corresponding initial intervention instruction set from a preset intervention strategy library according to the generated physiological risk markers; and dynamically calibrating parameters of the initial intervention instruction set by fusing the state deviation degree sequence, so as to finally generate an executable final intervention instruction set.

[0021] Embodiment 1: refer to Figure 2A multi-scale sliding window is applied to the heart rate signal stream to obtain a series of heart rate signal segments. Local extreme point detection is performed on each heart rate signal segment to determine the time coordinates of the start point, peak point and end point of each heart beat cycle. According to the determined time coordinates, the waveforms of all heart beat cycles are aligned and superimposed for average processing to obtain a reference waveform of heart rate rhythm. The reference waveform of heart rate rhythm is subtracted from each heart rate signal segment to obtain a heart rate residual signal. Power spectral density estimation is performed on the heart rate residual signal to calculate the energy proportion in the ultra-low frequency, low frequency and high frequency bands, forming a quantification spectrum of heart rate variability. Adaptive mode decomposition is applied to the respiratory signal stream to decompose it into a plurality of intrinsic mode function components and a residual trend component. The target component with energy concentrated near the respiratory frequency is identified from the intrinsic mode function components, and the target component is reconstructed into a pure respiratory signal. The residual trend component is subtracted from the original respiratory signal stream to obtain a respiratory corrected signal with slow trend removed. The instantaneous frequency of the pure respiratory signal is calculated to obtain a waveform of the change of respiratory fundamental frequency over time. Absolute value processing and low-pass filtering are performed on the respiratory corrected signal to obtain a change envelope line representing the slow change of respiratory depth.

[0022] Three-dimensional linear acceleration data and three-dimensional angular velocity data are separated from the posture motion signal stream. The three-dimensional linear acceleration data is twice integrated, and the three-dimensional angular velocity data is combined to obtain the displacement trajectory of the driver's body part in the three-dimensional space through attitude calculation. The projection components of the displacement trajectory on each axis of the spatial rectangular coordinate system are respectively subjected to Fourier transform to obtain the frequency spectrum in each axis direction, and the distribution spectrum of body motion energy in each axis direction is formed by these frequency spectrums. Significant peaks in the frequency spectrum in each axis direction are detected, and the frequencies corresponding to the peaks are compared with a preset typical body motion frequency range to screen out frequency points belonging to periodic motion to form a feature frequency set.

[0023] In specific implementation, a multi-scale sliding window is applied to the heart rate signal stream to obtain a series of heart rate signal segments, and in some embodiments, the length of the sliding window can be set to cover at least ten consecutive heart beat cycles. Local extreme point detection is performed on each heart rate signal segment to determine the time coordinates of the start point, peak point and end point of each heart beat cycle, and the local extreme point detection is realized by comparing the amplitude relationship between the signal point and its adjacent points. According to the determined time coordinates, the waveforms of all heart beat cycles are aligned and superimposed for average processing to obtain a reference waveform of heart rate rhythm, and the waveform alignment is subjected to time offset correction with the start point as the reference. The reference waveform of heart rate rhythm is subtracted from each heart rate signal segment to obtain a heart rate residual signal, and the heart rate residual signal mainly contains information of heart rate variability. Power spectral density estimation is performed on the heart rate residual signal to calculate the energy proportion in the ultra-low frequency, low frequency and high frequency bands, forming a quantification spectrum of heart rate variability, and the power spectral density estimation adopts the Welch method.

[0024] Adaptive mode decomposition (ADD) is applied to the respiratory signal stream, decomposing it into multiple intrinsic mode function (IMF) components and a residual trend component. ADD uses an iterative filtering process to decompose the signal into components of different scales. Target components with energy concentrated near the respiratory frequency are identified among the IMF components, and these target components are reconstructed into a clean respiratory signal. The energy concentration is determined by calculating the integral of the power spectrum of each component within a preset respiratory frequency band. The residual trend component is subtracted from the original respiratory signal stream to obtain a respiratory correction signal with the slow-varying trend removed; the residual trend component represents the slow baseline drift in the signal. The instantaneous frequency of the clean respiratory signal is calculated, yielding the waveform of the respiratory fundamental frequency changing over time. The instantaneous frequency is obtained through Hilbert transform and phase differentiation. The respiratory correction signal is then subjected to absolute value processing and low-pass filtering to obtain the envelope representing the slow changes in respiratory amplitude, with the cutoff frequency of the low-pass filter lower than the respiratory fundamental frequency.

[0025] Three-dimensional linear acceleration data and three-dimensional angular velocity data are separated from the attitude motion signal stream. These data are output from the accelerometer and gyroscope of the inertial measurement unit, respectively. The three-dimensional linear acceleration data is integrated twice, and combined with the three-dimensional angular velocity data, attitude calculation is performed to obtain the displacement trajectory of the driver's body parts in three-dimensional space. The attitude calculation uses complementary filtering or Kalman filtering algorithms to fuse multi-sensor data to compensate for integration errors. Fourier transforms are performed on the projection components of the displacement trajectory on each axis of the spatial rectangular coordinate system to obtain the spectrum for each axis. The distribution spectrum of body kinetic energy on each axis is composed of these spectra. In some embodiments, the spectrum for each axis is normalized for comparison. Significant peaks in the spectrum of each axis are detected, and the frequencies corresponding to these peaks are compared with a preset typical body motion frequency range. Frequency points belonging to periodic motion are selected to form a characteristic frequency set. It can be understood that the detection of significant peaks is achieved by setting an amplitude threshold, for example, satisfying the following relationship: , in: For frequency Spectral amplitude at that location This is the mean of the axial spectral amplitude. Standard deviation, This is the preset amplitude coefficient.

[0026] Example 2: See Figure 3The ultra-low frequency energy proportion, the low-to-high frequency energy ratio, and the total power value are extracted from the quantized spectrogram, and these values are used as a first group of feature vectors. The mean, variance, slope, and zero-crossing rate of the variation envelope are extracted from the variation envelope, and these values are used as a second group of feature vectors. The dominant frequency value, frequency concentration degree, and frequency stability coefficient are extracted from the characteristic frequency set, and these values are used as a third group of feature vectors. The first group of feature vectors, the second group of feature vectors, and the third group of feature vectors are spliced in a predetermined order to form a row vector. The above steps are repeated on a plurality of consecutive time frames, and the obtained plurality of row vectors are stacked in time order to form a multi-dimensional physiological feature matrix, wherein each row of the matrix corresponds to the multi-dimensional features of a time frame.

[0027] In a specific implementation, the ultra-low frequency energy proportion, the low-to-high frequency energy ratio, and the total power value are extracted from the quantized spectrogram, the low-to-high frequency energy ratio is the ratio of the low frequency band energy to the high frequency band energy, and the total power value is the integral of the frequency spectrum energy in a preset frequency band. These values are used as a first group of feature vectors, and it can be understood that the first group of feature vectors are used to represent the activity state of the autonomic nervous system. The mean of the variation envelope, the variance of the variation envelope, the slope of the variation envelope, and the zero-crossing rate of the variation envelope are extracted from the variation envelope, the slope of the variation envelope is obtained by linear fitting of the variation envelope, and the zero-crossing rate of the variation envelope is the ratio of the number of times the variation envelope crosses the mean level line to the time length. These values are used as a second group of feature vectors. The dominant frequency value, frequency concentration degree, and frequency stability coefficient are extracted from the characteristic frequency set, the dominant frequency value is the frequency point with the highest amplitude in the characteristic frequency set, the frequency concentration degree is used to quantify the degree of aggregation of frequency points around the dominant frequency value, and the frequency stability coefficient is the reciprocal of the standard deviation of the dominant frequency value in a plurality of consecutive time windows. These values are used as a third group of feature vectors.

[0028] The first group of feature vectors, the second group of feature vectors and the third group of feature vectors are spliced in a predetermined order to form a row vector, and in some embodiments, the predetermined order can be [ultra-low frequency energy proportion, low frequency to high frequency energy ratio, total power value, mean of variation envelope, variance of variation envelope, slope of variation envelope, zero-crossing rate of variation envelope, main frequency value, frequency concentration degree, frequency stability coefficient]. The steps of extracting the ultra-low frequency energy proportion, the low frequency to high frequency energy ratio and the total power value from the quantized spectrogram, extracting the mean, variance, slope and zero-crossing rate from the variation envelope, and extracting the main frequency value, the frequency concentration degree and the frequency stability coefficient from the characteristic frequency set are repeated on a plurality of continuous time frames. It can be understood that the length of each time frame is consistent with the length of the analysis window used to generate the quantized spectrogram, the variation envelope and the characteristic frequency set. The obtained plurality of row vectors are stacked in time sequence to form a multi-dimensional physiological feature matrix, wherein each row of the matrix corresponds to the multi-dimensional features of a time frame. In some embodiments, a row vector containing 10 features is stacked on 100 consecutive time frames to form a multi-dimensional physiological feature matrix of 100 rows and 10 columns.

[0029] The calculation of the frequency concentration degree can be realized by the form of information entropy, for example, quantified by the following relationship: , Wherein: represents the frequency concentration degree, represents the total number of frequency points in the characteristic frequency set, represents the normalized amplitude weight of the th frequency point, which is calculated by the proportion of the amplitude of the frequency point to the sum of the amplitudes of all points. Optionally, the smaller the value of the frequency concentration degree , the more concentrated the frequency energy is in a small number of frequency points. The calculation of the frequency stability coefficient is the reciprocal of the standard deviation of the main frequency value in a plurality of consecutive time windows. If the standard deviation is , the frequency stability coefficient is , and the calculation of the standard deviation is based on the sequence of main frequency values in a historical period of time.

[0030] Example 3: Obtain a preset driver state template, which includes a normal state feature template, a fatigue state feature template, and a disease risk state feature template. Calculate the Euclidean distance between each row of the multidimensional physiological feature matrix's feature vectors and the normal state feature template to form a basic deviation sequence. Calculate the similarity between each row of the multidimensional physiological feature matrix's feature vectors and the fatigue state feature template, and convert the similarity into a fatigue deviation. Calculate the matching score between each row of the multidimensional physiological feature matrix's feature vectors and a specific risk pattern in the disease risk state feature template. When the value of the basic deviation sequence continuously exceeds a warning threshold, an attention distraction risk label is generated. When the fatigue deviation is consistently higher than the fatigue threshold, a fatigue accumulation risk label is generated. When the matching score with a specific risk pattern exceeds a risk threshold, a sudden health risk label is generated.

[0031] In response to an attention distraction risk marker, an attention arousal strategy is retrieved from the intervention strategy library. This strategy includes initial parameters for acoustic stimulus type, visual cue content, and trigger duration. In response to a fatigue accumulation risk marker, an anti-fatigue strategy is retrieved from the intervention strategy library. This strategy includes initial parameters for suggested rest intervals, seat massage modes, and air conditioning adjustment targets. In response to a sudden health risk marker, an emergency response strategy is retrieved from the intervention strategy library. This strategy includes initial parameters for vehicle control suggestions, contact priorities, and information broadcast templates. The initial parameters of the selected strategies are then aggregated to form an initial intervention instruction set.

[0032] In practical implementation, a preset driver state template is obtained. This template includes a normal state feature template, a fatigue state feature template, and a disease risk state feature template. These templates are obtained through statistical learning of the physiological characteristics of corresponding states from a large amount of historical driving data. The Euclidean distance between each row of the multidimensional physiological feature matrix and the normal state feature template is calculated to form a basic deviation sequence. The formula for calculating the Euclidean distance is: , in: Represents the fundamental deviation of a single feature vector. This represents the eigenvector of the current row in the multidimensional physiological feature matrix. One element, The first element representing the feature template vector of the normal state One element, The total dimension number of the feature vector is represented. The similarity between each row feature vector of the multi-dimensional physiological feature matrix and the fatigue state feature template is calculated, and the similarity is converted into a fatigue deviation degree. It can be understood that the similarity calculation adopts cosine similarity, and the fatigue deviation degree is defined as the difference between 1 and the cosine similarity value. The matching score between each row feature vector of the multi-dimensional physiological feature matrix and a specific risk pattern in the disease risk state feature template is calculated. When the value of the basic deviation sequence continuously exceeds the early warning threshold, a distraction risk marker is generated. It can be understood that the judgment of continuous exceeding requires the basic deviation sequence to be greater than the early warning threshold continuously within a preset time window. When the fatigue deviation degree is continuously higher than the fatigue threshold, a fatigue accumulation risk marker is generated. In some embodiments, the judgment of continuous higher is based on the sliding average value of the fatigue deviation degree within a plurality of analysis periods being higher than the fatigue threshold. When the matching score with the specific risk pattern exceeds the risk threshold, a sudden health risk marker is generated, and the risk threshold is a pre-set high threshold.

[0033] In response to the distraction risk marker, an attention awakening strategy is called from the intervention strategy library, the attention awakening strategy contains initial parameters of acoustic stimulation type, visual prompt content and trigger duration, the initial parameters of the acoustic stimulation type include the tone and rhythm of the prompt sound, and the initial parameters of the visual prompt content include the text or icon content displayed on the dashboard. In response to the fatigue accumulation risk marker, an anti-fatigue strategy is called from the intervention strategy library, the anti-fatigue strategy contains initial parameters of recommended rest interval, seat massage mode and air conditioning adjustment target, the initial parameters of the seat massage mode specify the massage part and the basic intensity level, and the initial parameters of the air conditioning adjustment target specify the target temperature value and the air volume. In response to the sudden health risk marker, an emergency response strategy is called from the intervention strategy library, the emergency response strategy contains initial parameters of vehicle control suggestion, contact priority and information broadcast template, the initial parameters of the vehicle control suggestion contain “suggesting to slow down and park on the side”, the initial parameters of the contact priority specify the priority of contacting the emergency center, and the initial parameters of the information broadcast template contain pre-recorded voice warning content. The initial parameters of the acoustic stimulation type, the visual prompt content, the trigger duration, the recommended rest interval, the seat massage mode, the air conditioning adjustment target, the vehicle control suggestion, the contact priority and the information broadcast template of the selected strategy are summarized to form an initial intervention instruction set. Optionally, the initial intervention instruction set is transmitted internally in the system in the form of a structured data dictionary or a list.

[0034] Example 4: Read the latest values ​​and trends of the state deviation sequence. Based on the magnitude of the latest values, linearly adjust the intensity parameters of the acoustic stimuli or the frequency parameters of the visual cues in the initial intervention command set. Based on the slope of the trend, non-linearly adjust the duration parameters of the suggested rest intervals in the initial intervention command set. Combining the current time, historical driving time, and ambient light intensity, weighted corrections are made to the intensity curve of the seat massage mode and the specific temperature value of the air conditioning adjustment target. Integrate all parameters after adjustments for intensity, frequency, duration, and specific target values ​​to form the final intervention command set that can be directly parsed by the vehicle's execution system.

[0035] In practice, the latest values ​​and trends of the state deviation sequence are read. The trend is obtained by calculating the slope of a linear regression of the state deviation sequence within a recent time window. Based on the magnitude of the latest value, the intensity parameter of the acoustic stimulus or the frequency parameter of the visual cue in the initial intervention instruction set is linearly adjusted. It can be understood that the intensity parameter of the acoustic stimulus and the frequency parameter of the visual cue are directly proportional to the latest value. Based on the slope of the trend, the duration parameter of the suggested rest interval in the initial intervention instruction set is non-linearly adjusted. In some embodiments, when the duration parameter is... At that time, the final adjusted duration It can be calculated using the following formula: , in: This indicates the recommended rest interval duration after dynamic calibration. This parameter represents the duration of the suggested rest interval in the initial intervention instruction set. The slope representing the trend of the state deviation sequence. This is the proportionality coefficient. This is a non-linear factor used to control the sensitivity of the adjustment. The intensity curve of the seat massage mode and the specific temperature value of the air conditioning target are weighted and corrected by combining the current time, historical driving time, and ambient light intensity. The intensity curve correction of the seat massage mode increases the base intensity level as historical driving time increases, while the correction of the specific temperature value of the air conditioning target is based on a combination of the current time and ambient light intensity. It can be understood that the ambient light intensity is obtained through the vehicle's light sensor, and its value is used to fine-tune the target temperature of the air conditioning. In some embodiments, refer to Table 1 for the ambient light intensity and air conditioning temperature correction values.

[0036] Table 1: Mapping Table of Ambient Light Intensity and Air Conditioning Temperature Correction Values

[0037] Table 1 shows that during the daytime when the light is strong, the target temperature is inclined to be lowered to provide a cooler environment. The weighting correction process integrates the coefficients corresponding to the current time, the coefficients corresponding to the historical driving time, and the temperature correction value corresponding to the ambient light intensity obtained from Table 1 to obtain the final parameters of the seat massage mode intensity curve and the specific temperature value of the air conditioning adjustment target. After adjusting all parameters such as intensity, frequency, duration, and specific target value, the final intervention instruction set that can be directly parsed by the vehicle execution system is formed, and the final intervention instruction set is usually packaged and transmitted in a standardized vehicle network data frame format.

[0038] Referring to Figure 4 The figure is the core visualization result of the intervention instruction dynamic calibration link, focusing on the correlation between the state deviation slope and the recommended rest interval. The horizontal axis represents the state deviation slope, reflecting the deterioration rate of the driver's physiological state deviation from the normal mode; the vertical axis is the recommended rest interval, which is a key parameter in the anti-fatigue intervention strategy. This trend is consistent with the logic of adjusting the rest interval according to the nonlinear trend of the deviation, which not only ensures that the rest time is not excessively increased when the deterioration is slight, but also prolongs the rest interval when the state deteriorates rapidly, making the intervention strategy more adaptive to the real-time state of the driver, and embodying the individualization and gradient characteristics of the dynamic calibration mechanism.

[0039] In embodiment 5, the feedback physiological information set of the target driver after the implementation of the intervention measure is received in real time. The heart rate variability spectrum, the respiratory depth change envelope, and the body motion feature frequency set of the subsequent time period are extracted from the feedback physiological information set to form a feedback feature matrix. The feature drift between the feedback feature matrix and the multi-dimensional physiological feature matrix used to generate the final intervention instruction set is calculated. If the feature drift is greater than the adaptive update threshold, the corresponding state feature template in the preset driver state template is incrementally updated using the feedback feature matrix.

[0040] A piezoelectric sensing array is arranged below the surface of the vehicle seat to detect the original electrical signals of the heart rate signal stream and the respiratory signal stream caused by the slight pressure fluctuations due to heartbeats and respiration. Distributed inertial sensing units are embedded in the backrest and cushion of the vehicle seat to generate original motion data of the attitude motion signal stream by measuring the three-dimensional acceleration and angular velocity of the driver's torso and hips. High-precision analog-to-digital converters are used to synchronize sampling and digitize the original electrical signals and the original motion data, ensuring that the heart rate signal stream, the respiratory signal stream, and the attitude motion signal stream have a unified and synchronized time reference.

[0041] In a specific implementation, a set of feedback physiological information of the target driver after the implementation of the intervention measure is received in real time, the set of feedback physiological information is acquired in the same way as the set of original physiological information, and contains a heart rate signal stream, a respiration signal stream, and a posture motion signal stream in a subsequent time period. A heart rate variability quantization spectrum, a respiration depth change envelope, and a body motion characteristic frequency set in the subsequent time period are extracted from the set of feedback physiological information, the extraction method is the same as that for extracting the quantization spectrum, the change envelope, and the characteristic frequency set from the set of original physiological information, and a feedback feature matrix consistent with the dimension of the multi-dimensional physiological feature matrix is formed. A feature drift between the feedback feature matrix and the multi-dimensional physiological feature matrix on which the final intervention instruction set is generated is calculated, and in some embodiments, the feature drift is calculated by calculating the average Euclidean distance of the feature vectors of the corresponding rows of the two matrices: , wherein: represents the feature drift, represents the number of rows (the number of time frames) of the feature matrix, represents the dimension of the feature vector, represents the feature value of the first row and the first column of the feedback feature matrix, represents the feature value of the first row and the first column of the multi-dimensional physiological feature matrix on which the final intervention instruction set is generated. If the feature drift is greater than an adaptive update threshold, the feedback feature matrix is used to incrementally update the corresponding state feature template in the preset driver state template, and the incremental update adopts a sliding average or an exponential weighted average method. It can be understood that the update process only modifies the state feature template corresponding to the currently detected physiological risk marker in the preset driver state template.

[0042] A piezoelectric sensing array is arranged below the surface of the vehicle seat to detect the original electrical signals of the heart rate signal stream and the respiration signal stream caused by the slight pressure fluctuations due to heartbeats and respiration. The piezoelectric sensing array is distributed in a matrix form to facilitate the capture of physiological signals at different positions of the torso. Distributed inertial sensing units are embedded in the backrest and the cushion of the vehicle seat to generate original motion data of the posture motion signal stream by measuring the three-dimensional acceleration and angular velocity of the driver's torso and hips. Each distributed inertial sensing unit contains a three-axis accelerometer and a three-axis gyroscope. A high-precision analog-to-digital converter is used to synchronously sample and digitize the original electrical signals and the original motion data. In some embodiments, the high-precision analog-to-digital converter synchronously collects multiple analog signals at a sampling rate of no less than 1 kHz. Optionally, the synchronization is ensured by a unified hardware clock trigger, which ensures that the heart rate signal stream, the respiration signal stream, and the posture motion signal stream have a unified and synchronous time reference.

[0043] Referring to Figure 5 The figure is the core visualization result of the model adaptive update link, focusing on the change degree of physiological characteristics after intervention and the template update trigger logic. The horizontal axis is time, and the vertical axis is the feature drift amount. The dashed line represents the adaptive update threshold. The feature drift amount reflects the change amplitude of the driver's physiological characteristics after intervention. When the value exceeds the threshold, the system will automatically update the preset driver state template. The figure clearly presents the trigger timing of the template update, embodies the mechanism of dynamically optimizing the state template based on feedback physiological information, and realizes the adaptive iteration of the state template by monitoring the dynamic change of the feature drift amount, thereby improving the accuracy of subsequent state recognition.

[0044] It should be noted that, in this document, relational terms such as first and second and the like can be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Moreover, the terms "comprises", "comprising", or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but can include other elements not expressly listed or inherent to such process, method, article, or apparatus.

[0045] Although embodiments of the present application have been shown and described, it is to be understood that various modifications, substitutions, alternatives, and variations can be made in the embodiments without departing from the spirit and scope of the present application as defined by the appended claims and their equivalents.

Claims

1. A method for monitoring the state of a car driver based on the extraction of physiological signs, characterized in that, The method includes the following steps: Continuously capture the target driver's raw physiological information set, which includes heart rate signal stream, respiratory signal stream, and posture motion signal stream; The heart rate signal stream is subjected to time-domain morphological separation and frequency-domain spectral analysis to obtain the baseline waveform of the heart rate rhythm and the quantized spectrum of heart rate variability. Perform non-stationary signal decomposition and baseline drift elimination on the respiratory signal stream to extract the respiratory fundamental frequency waveform and the envelope of respiratory depth changes; The attitude motion signal stream is subjected to three-dimensional spatial trajectory reconstruction and frequency component division to obtain the distribution spectrum of body kinetic energy on each axis in space and the characteristic frequency set of periodic motion; A multidimensional physiological feature matrix is ​​constructed based on the quantized spectrum, the changing envelope, and the set of characteristic frequencies. The multidimensional physiological feature matrix is ​​matched and the difference is calculated layer by layer with the preset driver state template to generate a state deviation sequence and multiple physiological risk markers. Based on the physiological risk markers, select the corresponding initial intervention instruction set from the preset intervention strategy library; The parameters of the initial intervention instruction set are dynamically calibrated by integrating the state deviation sequence to generate the final intervention instruction set; The construction of a multidimensional physiological feature matrix based on the quantized spectrum, the changing envelope, and the set of feature frequencies includes: The ultra-low frequency energy ratio, the low frequency to high frequency energy ratio, and the total power value are extracted from the quantized spectrum and used as the first set of feature vectors. Extract the mean, variance, slope, and zero-crossing rate from the changing envelope to form the second set of feature vectors; Extract the main frequency value, frequency concentration, and frequency stability coefficient from the set of characteristic frequencies to form the third set of feature vectors; The first set of feature vectors, the second set of feature vectors, and the third set of feature vectors are concatenated in a predetermined order to form a row vector; Repeat the above steps over multiple consecutive time frames, and stack the resulting row vectors in chronological order to form the multidimensional physiological feature matrix, where each row of the matrix corresponds to the multidimensional feature of a time frame.

2. The method for monitoring the state of a car driver based on physiological signs extraction according to claim 1, characterized in that, The step of performing time-domain morphological separation and frequency-domain spectral analysis on the heart rate signal stream to obtain the baseline waveform of the heart rate rhythm and the quantized spectrum of heart rate variability includes: A multi-scale sliding window is applied to the heart rate signal stream to obtain a series of heart rate signal segments; For each heart rate signal segment, local extreme points are detected to determine the time coordinates of the start point, peak point, and end point of each heartbeat cycle; Based on the time coordinates, the waveforms of all heartbeat cycles are aligned, superimposed, and averaged to obtain the reference waveform of the heart rate rhythm. The heart rate residual signal is obtained by subtracting the reference waveform from each of the heart rate signal segments; The power spectral density of the heart rate residual signal is estimated, and its energy proportion in the ultra-low frequency, low frequency, and high frequency bands is calculated to form a quantitative spectrum of the heart rate variability.

3. The method for monitoring the state of a car driver based on physiological signs extraction according to claim 2, characterized in that, The step of performing non-stationary signal decomposition and baseline drift elimination on the respiratory signal stream, and extracting the change envelope of the respiratory fundamental frequency waveform and respiratory depth, includes: Adaptive mode decomposition is applied to the respiratory signal stream, decomposing it into multiple intrinsic mode function components and a residual trend component; Identify the target component in the intrinsic mode function components whose energy is concentrated near the respiratory rate, and reconstruct the target component into a pure respiratory signal; Subtracting the residual trend component from the respiratory signal stream yields a respiratory correction signal with the slow-varying trend removed; Calculate the instantaneous frequency of the pure respiratory signal to obtain the waveform of the respiratory fundamental frequency changing over time; The respiratory correction signal is subjected to absolute value processing and low-pass filtering to obtain the change envelope of the respiratory depth, which characterizes the slow change of respiratory amplitude.

4. The method for monitoring the state of a car driver based on physiological signs extraction according to claim 1, characterized in that, The process of performing three-dimensional spatial trajectory reconstruction and frequency component segmentation on the attitude motion signal stream yields the distribution spectrum of body kinetic energy along each spatial axis and the characteristic frequency set of periodic motion, including: Three-dimensional linear acceleration data and three-dimensional angular velocity data are separated from the attitude motion signal stream; The displacement trajectory of the driver's body parts in three-dimensional space is obtained by performing a second integral on the three-dimensional linear acceleration data and combining it with the three-dimensional angular velocity data through attitude calculation. The projection components of the displacement trajectory on each axis of the spatial rectangular coordinate system are subjected to Fourier transform to obtain the spectrum on each axis. The distribution spectrum of the body kinetic energy on each axis of space is composed of the spectrum. Significant peaks in the spectrum along each axis are detected, and the frequencies corresponding to the peaks are compared with a preset range of typical body motion frequencies to filter out frequency points belonging to periodic motion, thus forming the characteristic frequency set.

5. The method for monitoring the state of a car driver based on physiological signs extraction according to claim 1, characterized in that, The process involves performing layer-by-layer matching and difference calculation between the multidimensional physiological feature matrix and a preset driver state template to generate a state deviation sequence and multiple physiological risk markers, including: Obtain the preset driver status template, which includes a normal state feature template, a fatigue state feature template, and a disease risk state feature template. Calculate the Euclidean distance between each row of the feature vector of the multidimensional physiological feature matrix and the normal state feature template to form a basic deviation sequence; Calculate the similarity between each row of the feature vector of the multidimensional physiological feature matrix and the fatigue state feature template, and convert the similarity into fatigue deviation. Calculate the matching score between the feature vector of each row of the multidimensional physiological feature matrix and a specific risk pattern in the disease risk state feature template; When the value of the basic deviation sequence continuously exceeds the warning threshold, an attention distraction risk marker is generated; When the fatigue deviation is consistently higher than the fatigue threshold, a fatigue accumulation risk marker is generated. When the score matching a specific risk pattern exceeds a risk threshold, a sudden health risk marker is generated.

6. The method for monitoring the state of a car driver based on physiological signs extraction according to claim 5, characterized in that, The step of selecting a corresponding initial intervention instruction set from a preset intervention strategy library based on the physiological risk markers includes: In response to the attention distraction risk marker, an attention reawakening strategy is invoked from the intervention strategy library. The attention reawakening strategy includes initial parameters for acoustic stimulus type, visual cue content, and trigger duration. In response to the fatigue accumulation risk marker, an anti-fatigue strategy is invoked from the intervention strategy library, the anti-fatigue strategy including initial parameters for suggested rest intervals, seat massage modes, and air conditioning adjustment targets; In response to the sudden health risk marker, an emergency response strategy is invoked from the intervention strategy library. The emergency response strategy includes initial parameters for vehicle control recommendations, contact priorities, and information broadcast templates. The initial parameters of the selected strategy are summarized to form the initial intervention instruction set.

7. The method for monitoring the state of a car driver based on physiological signs extraction according to claim 6, characterized in that, The parameters of the initial intervention instruction set are dynamically calibrated by fusing the state deviation sequence to generate the final intervention instruction set, including: Read the latest values ​​and trends of the state deviation sequence; Based on the magnitude of the latest value, the intensity parameter of the acoustic stimulus or the frequency parameter of the visual cue in the initial intervention instruction set are linearly adjusted; Based on the slope of the changing trend, the duration parameter of the rest interval suggested in the initial intervention instruction set is adjusted non-linearly. The intensity curve of the seat massage mode and the specific temperature value of the air conditioning adjustment target are weighted and corrected by combining the current time, historical driving time and ambient light intensity. All parameters, after being adjusted for intensity, frequency, duration, and specific target values, are integrated to form a final set of intervention commands that can be directly parsed by the vehicle's execution system.

8. The method for monitoring the state of a car driver based on physiological signs extraction according to claim 1, characterized in that, Following the generation of the final intervention instruction set, a model adaptive update step is also included: Real-time reception of the target driver's feedback physiological information set after the intervention measures are implemented; The heart rate variability quantification spectrum, respiratory depth change envelope and body movement characteristic frequency set for subsequent time periods are extracted from the feedback physiological information set to form a feedback feature matrix. Calculate the feature drift between the feedback feature matrix and the multidimensional physiological feature matrix on which the final intervention instruction set was generated; If the feature drift is greater than the adaptive update threshold, the feedback feature matrix is ​​used to incrementally update the corresponding state feature template in the preset driver state template.

9. The method for monitoring the state of a car driver based on physiological signs extraction according to claim 1, characterized in that, The continuously captured set of raw physiological information of the target driver includes: A piezoelectric sensor array is deployed beneath the surface of the vehicle seat to generate the raw electrical signals of the heart rate signal stream and the respiratory signal stream by detecting minute pressure fluctuations caused by heartbeats and breathing. Distributed inertial sensing units are embedded in the backrest and cushion of the vehicle seat to generate the raw motion data of the attitude motion signal stream by measuring the three-dimensional acceleration and angular velocity of the driver's torso and hips. A high-precision analog-to-digital converter is used to synchronously sample and digitize the original electrical signals and the original motion data to ensure that the heart rate signal stream, the respiratory signal stream, and the posture motion signal stream have a unified and synchronized time reference.

Images