An adaptive adjustment system of a car massage seat

Abstract

The application provides an adaptive adjustment system of a car massage seat, and relates to the technical field of intelligent comfort control of car interiors, and comprises: a multi-source sensing acquisition module, which is used for acquiring a body pressure distribution signal in a seat surface, a road surface vibration signal transmitted by a vehicle chassis, a deformation signal of a seat framework, a muscle tension signal and a heart rate signal of a driver in real time; a fusion decoupling module, which is used for jointly processing the body pressure distribution signal, the road surface vibration signal, the deformation signal and the muscle tension signal, separating body pressure coupling interference caused by road unevenness and sitting posture change, obtaining real-time sitting posture angle, road surface excitation intensity level and muscle group tension distribution of the driver, and forming compensated body pressure distribution change information. The application can solve the problems of distorted dynamic driving condition sensing perception data, low massage point position and seat support adjustment matching degree and poor fatigue relief effect, and effectively improves the driving comfort in the whole process of driving in all road conditions.

Description

Technical Field

[0001] This invention relates to the field of intelligent comfort control technology for automotive interiors, and in particular to an adaptive adjustment system for an automotive massage seat. Background Technology

[0002] Nowadays, in-car massage seats are widely adapted to all driving scenarios, such as long-distance commuting, cruising on bumpy roads in the suburbs, and long-term idling. They can effectively relieve the tension and discomfort of the back, shoulders and neck muscles of drivers and passengers from sitting for a long time, and alleviate driving fatigue.

[0003] However, most of the mainstream intelligent massage seat control solutions currently on the market have the following technical defects: it is difficult to isolate in real time the cross-interference of body pressure caused by the coupling of the impact of uneven road surface during driving and the dynamic sitting posture deviation of the driver and passengers. For example, when a vehicle is traveling on bumpy roads or roads with continuous potholes, the chassis continuously transmits high-frequency vibrations that disturb the entire force distribution on the seat. This is compounded by the driver's constant dynamic changes in posture, such as adjusting the backrest angle, shifting the body to control the vehicle, and slightly shifting the buttocks. The combination of these two conditions creates body pressure coupling noise interference. Existing systems struggle to accurately decouple and remove invalid interference signals, leading to distortion of body pressure and muscle group data. This makes it impossible to accurately determine the driver's actual fatigue points in the lower back, shoulders, and neck, as well as the core areas of muscle tension. Furthermore, the massage points, intensity, frequency, and lumbar support and side wing adjustment parameters output based on distorted data deviate significantly from the actual comfort needs of drivers and passengers. On the one hand, this results in issues such as massage point misalignment, excessive localized pressure, and lack of massage in key fatigue areas. On the other hand, the seat's firmness and angle of support cannot match the real-time driving dynamics. This may result in the loss of the basic functions of massage to relieve fatigue and provide reasonable support, while also exacerbating localized pressure on the driver's torso, rapidly accumulating driving fatigue, and weakening the comfort of long-distance driving. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to provide an adaptive adjustment system for a car massage seat, which can effectively solve the problems of distortion of driving and riding sensation data, low matching degree between massage points and seat support adjustment, and poor fatigue relief effect under dynamic driving conditions, and effectively improve driving and riding comfort and driving and riding control adaptation safety throughout the entire driving process under all road conditions.

[0005] To solve the above-mentioned technical problems, the technical solution of the present invention is as follows:

[0006] Firstly, an adaptive adjustment system for a car massage seat includes:

[0007] The multi-source sensor acquisition module is used to acquire in real time the body pressure distribution signal in the seat surface, the road vibration signal transmitted by the vehicle chassis, the deformation signal of the seat frame, the driver's muscle tension signal, and the heart rate signal.

[0008] The fusion decoupling module is used to jointly process the body pressure distribution signal, road vibration signal, deformation signal and muscle tension signal, separate the body pressure coupling interference caused by road unevenness and sitting posture changes, and obtain the driver's real-time sitting posture angle, road excitation intensity level and muscle tension distribution to form compensated body pressure distribution change information.

[0009] The fatigue quantification module is used to obtain heart rate variability based on the heart rate signal, and calculate the real-time fatigue index based on the heart rate variability, combined with the compensated body pressure distribution change information and the muscle tension distribution.

[0010] The adaptive decision-making module is used to obtain the massage action sequence and seat support adjustment amount based on the real-time fatigue index, real-time sitting posture angle and road excitation intensity level; wherein the massage action sequence includes a massage point sequence that dynamically slides along the driver's spine, massage intensity value, frequency value and technique mode; the seat support adjustment amount includes lumbar support protrusion amount, side wing clamping angle and backrest tilt fine adjustment angle;

[0011] The execution control module is used to control the massage unit in the seat to perform massage according to the massage action sequence, and simultaneously drive the lumbar support adjustment mechanism, side wing clamping mechanism and backrest adjustment mechanism of the car massage seat to perform adjustment according to the seat support adjustment amount.

[0012] Furthermore, real-time acquisition of body pressure distribution signals within the seat surface, road vibration signals transmitted from the vehicle chassis, deformation signals of the seat frame, driver muscle tension signals, and heart rate signals includes:

[0013] By collecting the pressure distribution caused by body contact through multiple preset sensing points on the seat surface corresponding to the driver's lumbar spine, thoracic spine, sacrum and ischial tuberosity, a body pressure distribution signal is formed.

[0014] Mechanical vibrations caused by uneven road surfaces are picked up from the connection point between the seat frame and the vehicle chassis. The mechanical vibrations are converted into acceleration signals. The acceleration signals are then subjected to bandpass filtering and root mean square calculation to obtain the road vibration signal.

[0015] Detect minute deformations of the seat frame under the driver's weight and dynamic loads, and generate deformation signals;

[0016] Electromyographic activity information of target muscle groups in the driver's lower back and neck is obtained, and the electromyographic activity information is processed by full-wave rectification and moving average filtering to obtain muscle tension signal;

[0017] Heart rate-related physiological electrical signals or photoplethysmography signals are collected from the driver's body parts. The heart rate signal is obtained by detecting and extracting the heartbeat interval from the physiological electrical signals or the pulse interval from the photoplethysmography signals and converting them.

[0018] Furthermore, the body pressure distribution signal, road vibration signal, deformation signal, and muscle tension signal are jointly processed to separate the body pressure coupling interference caused by road unevenness and changes in sitting posture, obtaining the driver's real-time sitting posture angle, road excitation intensity level, and muscle tension distribution, forming compensated body pressure distribution change information, including:

[0019] The pressure center coordinates, contact area, and pressure values ​​at each preset sensing point are extracted from the body pressure distribution signal to form initial body pressure characteristics; the bending angle and torsion of the seat frame are extracted from the deformation signal to obtain seat deformation characteristics; the road vibration signal is subjected to spectral analysis to extract the dominant frequency and amplitude to obtain vibration characteristics; the root mean square of electromyographic amplitude is extracted from the muscle tension signal to obtain muscle group activation characteristics.

[0020] The vibration characteristics are correlated with the seat deformation characteristics, including separating the seat frame response components caused by road excitation, and evaluating the road excitation intensity level based on the dominant frequency and amplitude.

[0021] Based on the road excitation intensity level and seat deformation characteristics, the relative tilt angle between the driver's waist and the backrest and the relative rotation angle between the buttocks and the seat surface are calculated to obtain the real-time sitting posture angle.

[0022] Based on the real-time sitting angle, the muscle activation characteristics are normalized and the static tension component of the muscle group is corrected to obtain the muscle tension distribution.

[0023] Using the road surface excitation intensity level and real-time sitting posture angle as constraints, the pressure fluctuations synchronized with road surface vibration and the pressure offsets related to sitting posture tilt are extracted from the initial body pressure characteristics to form compensated body pressure distribution change information.

[0024] Furthermore, heart rate variability is obtained based on the heart rate signal. Based on the heart rate variability, combined with the compensated body pressure distribution change information and the muscle tension distribution, a real-time fatigue index is calculated, including:

[0025] Extract continuous heartbeat intervals from the heart rate signal, and perform time-domain or frequency-domain analysis on the continuous heartbeat intervals to obtain heart rate variability characteristics.

[0026] The real-time fatigue index is obtained by fusing the heart rate variability characteristics, the compensated body pressure distribution change information, and the muscle tension distribution.

[0027] Furthermore, based on the real-time fatigue index, real-time sitting angle, and road surface excitation intensity level, the massage action sequence and seat support adjustment amount are obtained, including:

[0028] Based on the real-time fatigue index and real-time sitting angle, basic comfort requirement parameters are obtained; among which, basic comfort requirement parameters include massage point sequence, massage technique mode, massage intensity base value and frequency base value;

[0029] The basic comfort requirement parameters are fused and compensated with the road surface excitation intensity level to obtain the compensated massage action sequence;

[0030] Based on the compensated massage action sequence and the real-time sitting posture angle, the seat support adjustment amount is calculated, which includes the lumbar support protrusion amount, the side wing clamping angle, and the backrest tilt fine adjustment angle.

[0031] Furthermore, based on the compensated massage motion sequence and the real-time sitting posture angle, the seat support adjustment amount is calculated. This seat support adjustment amount includes the lumbar support protrusion amount, the side wing clamping angle, and the backrest tilt fine-tuning angle, including:

[0032] The massage point sequence and technique pattern are extracted from the compensated massage action sequence. The target area for spinal support is determined based on the massage point sequence, and the support method requirement is determined based on the technique pattern.

[0033] Using the target area and support method requirements as input, and combined with the real-time sitting angle, the amount of lumbar support protrusion within the target area is calculated to obtain the lumbar support protrusion.

[0034] Based on the lumbar support protrusion and the real-time sitting angle, the clamping angle required to maintain the lateral stability of the driver's torso is calculated, and the side wing clamping angle is obtained.

[0035] Based on the lumbar support protrusion, the side wing clamping angle, and the real-time sitting posture angle, the backrest tilt adjustment angle is calculated to obtain the backrest tilt adjustment angle.

[0036] Furthermore, the step of controlling the massage unit within the seat to perform massage according to the massage action sequence, and simultaneously driving the lumbar support adjustment mechanism, side wing clamping mechanism, and backrest adjustment mechanism of the car massage seat to perform adjustments according to the seat support adjustment amount, includes:

[0037] According to the massage point sequence, massage intensity value, frequency value and technique mode in the compensated massage action sequence, the massage unit in the seat is driven to perform massage along the driver's spine, forming a massage execution feedback signal;

[0038] Based on the massage execution feedback signal and the lumbar support protrusion amount, the lumbar support adjustment mechanism is driven to perform lumbar support protrusion adjustment to obtain the lumbar support adjustment completed state.

[0039] Based on the lumbar support adjustment completion state and the side wing clamping angle, drive the side wing clamping mechanism to perform side wing clamping, and obtain the side wing adjustment completion state;

[0040] Based on the completion status of the side wing adjustment and the backrest pitch fine-tuning angle, the backrest adjustment mechanism is driven to perform backrest pitch fine-tuning.

[0041] In a second aspect, a computing device includes:

[0042] One or more processors;

[0043] A storage device for storing one or more programs that, when executed by one or more processors, cause the one or more processors to execute the system.

[0044] Thirdly, a computer-readable storage medium storing a program that, when executed by a processor, performs the system.

[0045] The above-described solution of the present invention has at least the following beneficial effects:

[0046] Because it employs a multi-source sensor acquisition module to fuse five types of signals in real time—body pressure distribution, road vibration, seat deformation, muscle tension, and heart rate—and uses a fusion decoupling module to separate body pressure coupling interference caused by road unevenness and changes in sitting posture, it generates compensated body pressure distribution change information. Simultaneously, it acquires real-time sitting posture angle, road excitation intensity level, and muscle tension distribution. This overcomes the technical problems of distorted single-dimensional sensing data and difficulty in removing body pressure coupling clutter under dynamic driving conditions. Furthermore, a fatigue quantification module calculates a real-time fatigue index by integrating heart rate variability, compensated body pressure changes, and muscle tension. An adaptive decision-making module dynamically generates massage points, intensity, frequency, techniques, and seat support adjustments based on the fatigue index, sitting posture angle, and road condition level. An execution control module drives the massage unit in coordination with the lumbar support, side wings, and backrest adjustment mechanisms, overcoming the poor adaptability of fixed preset modes. Ultimately, it achieves precise massage, coordinated support, and efficient fatigue relief under various road conditions and sitting postures, effectively improving long-distance driving comfort and handling safety. Attached Figure Description

[0047] Figure 1 This is a schematic diagram of an adaptive adjustment system for a car massage seat provided in an embodiment of the present invention.

[0048] Figure 2 This is a schematic diagram illustrating the process of an adaptive adjustment system for a car massage seat according to an embodiment of the present invention, which performs joint processing on the body pressure distribution signal, road vibration signal, deformation signal, and muscle tension signal to separate the body pressure coupling interference caused by road unevenness and changes in sitting posture. Detailed Implementation

[0049] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.

[0050] like Figure 1 As shown, an embodiment of the present invention proposes an adaptive adjustment system for a car massage seat, comprising:

[0051] The multi-source sensor acquisition module is used to acquire in real time the body pressure distribution signal in the seat surface, the road vibration signal transmitted by the vehicle chassis, the deformation signal of the seat frame, the driver's muscle tension signal, and the heart rate signal.

[0052] The fusion decoupling module is used to jointly process the body pressure distribution signal, road vibration signal, deformation signal and muscle tension signal, separate the body pressure coupling interference caused by road unevenness and sitting posture changes, and obtain the driver's real-time sitting posture angle, road excitation intensity level and muscle tension distribution to form compensated body pressure distribution change information.

[0053] The fatigue quantification module is used to obtain heart rate variability based on the heart rate signal, and calculate the real-time fatigue index based on the heart rate variability, combined with the compensated body pressure distribution change information and the muscle tension distribution.

[0054] The adaptive decision-making module is used to obtain the massage action sequence and seat support adjustment amount based on the real-time fatigue index, real-time sitting posture angle and road excitation intensity level; wherein the massage action sequence includes a massage point sequence that dynamically slides along the driver's spine, massage intensity value, frequency value and technique mode; the seat support adjustment amount includes lumbar support protrusion amount, side wing clamping angle and backrest tilt fine adjustment angle;

[0055] The execution control module is used to control the massage unit in the seat to perform massage according to the massage action sequence, and simultaneously drive the lumbar support adjustment mechanism, side wing clamping mechanism and backrest adjustment mechanism of the car massage seat to perform adjustment according to the seat support adjustment amount.

[0056] In this embodiment of the invention, by incorporating a multi-source sensor acquisition module, multiple driving condition and human physiological state signals can be collected simultaneously. A fusion and decoupling module accurately eliminates interference from road vibrations and changes in posture caused by body pressure coupling, effectively improving the accuracy of seat force perception data. Based on heart rate variability combined with body pressure and muscle group status, a real-time fatigue index is quantified, making fatigue assessment results more consistent with the driver's actual physical sensations. An adaptive decision-making module can generate dynamic massage action sequences that conform to the spine distribution and multi-dimensional seat support adjustments matching the current posture as needed. The entire adaptive adjustment system achieves synchronous closed-loop linkage adjustment of massage actions and seat support posture, effectively avoiding problems such as massage point misalignment and insufficient support fit. It reasonably and appropriately alleviates fatigue in the lower back, shoulders, and neck muscles during long-distance driving, optimizes driving posture support, and improves driving comfort and posture stability under complex road conditions.

[0057] In a preferred embodiment of the present invention, real-time acquisition of body pressure distribution signals within the seat surface, road vibration signals transmitted by the vehicle chassis, deformation signals of the seat frame, muscle tension signals of the driver, and heart rate signals includes:

[0058] By collecting pressure distribution data generated by body contact through multiple preset sensing points on the seat surface corresponding to the driver's lumbar, thoracic, sacral, and ischial tuberosities, a body pressure distribution signal is formed. Specifically, this includes: pre-arranging a flexible pressure sensor array on the seat surface, the flexible pressure sensor array containing multiple sensing points, which are arranged in zones according to the physiological curvature of the human spine and the support characteristics of the pelvis; wherein, the first sensing point group is located in the lower part of the seat back, corresponding to the L1 to L5 segments of the driver's lumbar spine, used to sense the contact pressure changes between the lumbar spine and the backrest in the forward lordosis state; the second sensing point group is located in the middle of the seat back, corresponding to the T1 to T12 segments of the driver's thoracic spine, used to sense the contact pressure distribution in the upper back leaning state and the scapular region; the third sensing point group is located at the junction of the lower part of the seat back and the rear part of the seat surface, corresponding to the driver's sacral region, used to sense the pressure changes in the sacral tilt and pelvic rotation states; the fourth sensing point group is located in the middle and rear part of the seat surface, corresponding to the area below the ischial tuberosities on both sides of the driver, used to sense the ischial weight-bearing state and the pelvic tilt trend.

[0059] In this system, each sensing point within a group is evenly arranged according to a set row and column spacing and embedded between the seat foam layer and the seat cover, so that it can directly bear the pressure transmitted by the torso after the driver sits down. Each sensing point synchronously samples pressure at a preset sampling frequency of 50Hz. This sampling frequency is higher than the typical frequency of human posture adjustment movements, and can capture transient pressure fluctuations caused by changes in posture in real time. After each sampling, the current pressure value of each sensing point is obtained, and according to the row and column coordinates set for each sensing point in the seat surface, the pressure value is filled into the corresponding row and column positions in a two-dimensional matrix corresponding to the spatial position of the seat surface, forming a pressure distribution matrix at the current sampling time. This serves as a body pressure distribution signal reflecting the weight distribution characteristics and contact morphology characteristics of the driver's torso.

[0060] The mechanical vibration caused by uneven road surface is picked up at the connection between the seat frame and the vehicle chassis. The mechanical vibration is converted into an acceleration signal, and the acceleration signal is bandpass filtered and root mean square calculated to obtain the road vibration signal. Specifically, a triaxial accelerometer is fixed at the connection between the seat frame and the vehicle chassis. The sensitive axes of the triaxial accelerometer are respectively aligned with the vertical, longitudinal, and lateral axes of the vehicle coordinate system. The vertical sensitive axis is mainly used to pick up the vertical mechanical vibration caused by uneven road surface. When the vehicle is moving, the road surface unevenness excitation is transmitted to the seat frame connection through the tires, suspension, and chassis. The piezoelectric or capacitive sensitive element inside the triaxial accelerometer generates a charge or voltage change proportional to the vibration acceleration under the action of inertial force, thereby converting the mechanical vibration borne by the seat frame connection into three-channel acceleration electrical signals in real time.

[0061] The vertical acceleration signal output from the triaxial accelerometer is bandpass filtered. The lower cutoff frequency of the bandpass filter is set to 0.5Hz to filter out ultra-low frequency components caused by sensor zero drift and vehicle steady-state tilt. The upper cutoff frequency is set to 80Hz to filter out high-frequency noise and structural local modal components unrelated to road surface excitation. The retained frequency band from 0.5Hz to 80Hz covers the main frequency range of vibration response on the seat frame caused by typical road surface irregularities during vehicle operation. The root mean square (RMS) of the filtered vertical acceleration signal is calculated segment by segment according to a preset time window. The calculation window length is set to 1 second to match the integral time constant of human vibration perception. The RMS calculation is performed according to the following formula: ;

[0062] in, This represents the root mean square value of acceleration within the time window. This represents the value of the filtered vertical acceleration signal at each sampling moment within the current time window; N represents the total number of sampling points of the filtered vertical acceleration signal within the current time window, i.e., the number of acceleration sampling data points participating in this root mean square calculation; the root mean square value of acceleration calculated by the above formula within each time window is used as the effective value of road vibration at that moment, thereby characterizing the road vibration signal of the current road vibration energy; the road vibration signal can reflect the excitation intensity of road unevenness transmitted to the seat surface through the chassis and seat frame, providing an accurate input basis for subsequent evaluation of the road excitation intensity level.

[0063] The system detects minute deformations of the seat frame under the driver's weight and dynamic loads, generating deformation signals. Specifically, it involves attaching multiple strain sensors to key stress-bearing areas of the seat frame. These key stress-bearing areas include the middle sections of the longitudinal beams on both sides of the backrest frame, the front and rear sections of the longitudinal beams on both sides of the seat frame, and the pivot bracket area at the connection between the backrest and the seat. The strain sensor in the middle section of the longitudinal beam of the backrest frame is used to detect bending deformation caused by the driver's lumbar support and lateral tilt. The strain sensors in the front and rear sections of the longitudinal beams of the seat frame are used to detect bending and torsional deformation caused by the driver's hip weight distribution and forward and backward adjustments to the sitting posture. The strain sensor in the pivot bracket area is used to detect shear deformation caused by changes in the angle between the backrest and the seat.

[0064] Each strain sensor employs a metal foil resistance strain gauge, which is attached to the frame surface in a full-bridge circuit configuration to compensate for the impact of temperature changes on measurement accuracy. When the driver's weight and dynamic loads such as vehicle acceleration, deceleration, and steering are applied to the seat frame, micro-strain is generated on the frame surface, and the resistance value of the strain gauge changes linearly accordingly. The full-bridge circuit converts this resistance change into a differential voltage signal. This differential voltage signal is then amplified by an instrumentation amplifier, filtered by a low-pass filter to remove high-frequency vibration noise and power frequency interference above 50Hz, and then converted into a digital signal by an analog-to-digital converter at a sampling frequency of 200Hz. This generates the deformation signal that reflects the structural response of the seat frame caused by changes in the driver's posture and the vehicle's motion.

[0065] Electromyographic (EMG) activity information of target muscle groups in the driver's lower back and neck is acquired. This EMG activity information is then processed using full-wave rectification and moving average filtering to obtain muscle tension signals. Specifically, this involves acquiring muscle activity information using EMG sensors attached to the surface of the target muscle groups in the driver's lower back and neck. These target muscle groups include the lumbar segment of the erector spinae, the upper trapezius, and the levator scapulae. Specifically, the EMG sensor for the lumbar segment of the erector spinae is attached 2 cm lateral to the spinous processes of L3 to L5 to detect muscle activation during lumbar support and forward tilting. The EMG sensor for the upper trapezius is attached at the midpoint of the line connecting the C7 spinous process to the acromion to detect shoulder shrugging and neck and shoulder tension. The EMG sensor for the levator scapulae is attached between the superior angle of the scapula and the transverse processes of C1 to C4 to detect muscle tension during lateral flexion and rotation of the neck.

[0066] In this process, each electromyography (EMG) sensor synchronously acquires weak bioelectrical signals generated by muscle fiber contraction at a preset sampling rate of 1000Hz. The amplitude of these weak bioelectrical signals is typically in the microvolt range. The acquired EMG activity information is first subjected to full-wave rectification, which takes the absolute value of the alternating positive and negative original EMG signals, so that the entire signal waveform is converted into a positive value, which facilitates the subsequent extraction of the muscle activity intensity envelope. Then, the rectified signal is subjected to moving average filtering, and the average value of the signal within the window is calculated point by point with a time window width of 100 milliseconds to remove high-frequency noise and smooth the signal contour, thereby extracting the envelope features of muscle electrical activity and obtaining the muscle tension signal that can reflect the real-time tension of the target muscle group.

[0067] The process involves collecting heart rate-related physiological electrical signals or photoplethysmography (PPG) signals from the driver's body, extracting the heartbeat interval from the physiological electrical signals, or extracting the pulse interval from the PPG signals, and converting these signals to obtain the heart rate signal. Specifically, this includes: collecting physiological electrical signals in the form of electrocardiograms (ECGs) from the driver's body, or collecting pulse wave signals in the form of PPGs (for example, ECGs can be collected from the driver's body via a wearable chest strap or steering wheel embedded electrodes, or PPGs can be collected via a finger clip or ear clip photoelectric sensor). If ECGs are used, the signal is bandpass filtered to extract QRS complex features in the 0.5Hz to 40Hz frequency band. Then, the R-wave peaks of each cardiac cycle are identified and located, and the time interval between two consecutive R-wave peaks is recorded as the heartbeat interval.

[0068] If a pulse wave signal in the form of photoplethysmography is used, the signal is bandpass filtered to extract the pulse wave characteristics in the frequency band from 0.5Hz to 5Hz. Then, the first derivative of the filtered waveform is calculated, and the position of the maximum value of the derivative is detected as the peak reference point. The time interval between two consecutive peaks is recorded as the pulse interval. The multiple heartbeat intervals or pulse intervals obtained consecutively are averaged, and 60 is divided by the average interval in seconds to convert it into the number of heartbeats per minute, thus obtaining the heart rate signal that can reflect the driver's current physiological load level.

[0069] like Figure 2 As shown, in another preferred embodiment of the present invention, the body pressure distribution signal, road vibration signal, deformation signal, and muscle tension signal are jointly processed to separate the body pressure coupling interference caused by road unevenness and changes in sitting posture, thereby obtaining the driver's real-time sitting posture angle, road excitation intensity level, and muscle tension distribution, forming compensated body pressure distribution change information, including:

[0070] The initial body pressure characteristics are formed by extracting the pressure center coordinates, contact area, and pressure values ​​at each preset sensing point from the body pressure distribution signal; the seat deformation characteristics are obtained by extracting the bending angle and torsion of the seat frame from the deformation signal; the vibration characteristics are obtained by performing spectral analysis on the road vibration signal and extracting the dominant frequency and amplitude; and the muscle activation characteristics are obtained by extracting the root mean square amplitude of the electromyography signal from the muscle tension signal. Specifically, this includes: extracting the pressure center coordinates, contact area, and pressure values ​​at each preset sensing point from the body pressure distribution signal to form the initial body pressure characteristics; wherein, the pressure center coordinates are calculated according to the following formula: In the formula, For the first The pressure value at each preset sensing point is expressed in Newtons. The preset row and column coordinates of the sensing point within the seat surface, in meters; The total number of sensing points; contact area The area of ​​the polygon enclosed by all sensing points with pressure values ​​greater than zero is represented in square meters; the pressure value at each sensing point is... It is also a component of the initial body pressure characteristics.

[0071] The bending angle and torsion of the seat frame are extracted from the deformation signal to obtain the seat deformation characteristics; bending angle The angular displacement sensor at the connection between the backrest and the seat directly outputs the value in degrees; the torque is also measured. The value is obtained by multiplying the strain difference between the longitudinal beams on both sides of the seat frame by a calibration coefficient, with the unit being degrees; spectral analysis of the road vibration signal is performed, and the dominant frequency is extracted using Fast Fourier Transform. With amplitude Vibration characteristics were obtained; among them The frequency at which the power spectral density reaches its maximum value; Corresponding to the vibrational energy at this frequency; the root mean square of the electromyographic amplitude is extracted from the muscle tension signal and calculated window by window according to the following formula: ;in, This represents the root mean square of the electromyographic amplitude within the current time window, in microvolts. For the first in the window The amplitude of electromyographic signals at each sampling point This represents the total number of sampling points within the window; the calculated target muscle groups (lumbar segment of erector spinae, upper trapezius, levator scapulae) Values ​​constitute the muscle group activation characteristics.

[0072] The vibration characteristics are correlated with the seat deformation characteristics, including separating the seat frame response component caused by road excitation and assessing the road excitation intensity level based on the dominant frequency and amplitude. Specifically, this includes: correlating the vibration characteristics with the seat deformation characteristics, separating the seat frame response component caused by road excitation, and assessing the road excitation intensity level based on the dominant frequency in the vibration characteristics. With amplitude The road surface excitation intensity level is assessed; specifically, the following vibration transfer function model is established: In the formula, For frequency variables; Let be the power spectral density of the road vibration signal, which is defined as the Fourier transform of the autocorrelation function of the road vibration acceleration signal, with units of square meters per second squared per hertz. The cross-power spectral density of the road vibration signal and the seat deformation signal is defined as the Fourier transform of their cross-correlation function, with units of degrees multiplied by meters per second squared per hertz. Let be the vibration transfer function of the system, representing the vibration transmission characteristics from the chassis to the seat frame; using the transfer function, the seat deformation response components caused by road vibration can be calculated, and the calculation formula is as follows: ;in, The Fourier spectrum of the road surface vibration acceleration signal. The Fourier spectrum of the corresponding seat deformation response components; for Performing an inverse Fourier transform yields the elastic deformation angle in the time domain. The unit is degrees; further, from the total bending angle in the seat deformation characteristics Subtract The remaining portion is attributed to the driver's active adjustment of their seating position; based on this, according to the dominant frequency in the vibration characteristics... and amplitude Assess the pavement excitation strength level Pre-set low threshold and high threshold ,when hour, (Low level); when hour, (Medium level); when hour, (High-level), among which and The units are all meters per second squared, determined through statistical analysis of a large amount of measured data from typical road surfaces.

[0073] Based on the road surface excitation intensity level and seat deformation characteristics, the relative tilt angle between the driver's waist and the backrest and the relative rotation angle between the buttocks and the seat surface are calculated to obtain the real-time sitting posture angle. Specifically, this includes: based on the obtained road surface excitation intensity level... Based on the obtained seat deformation characteristics, the relative inclination angle between the driver's waist and the backrest and the relative rotation angle between the buttocks and the seat surface are calculated to obtain the real-time sitting posture angle; among which, the relative inclination angle between the waist and the backrest... Calculated using the following formula: The The total bending angle of the backrest frame in the seat deformation characteristics; It is based on the road surface excitation strength level The elastic bending compensation angle, in degrees, is interpolated from a pre-calibrated road excitation-deformation response curve. This calibration curve was obtained through real-vehicle testing under both unloaded and loaded conditions on different road surfaces. Further details include the relative rotation angle between the hip and seat surface. The difference in torsional values ​​between the deformation sensors on the left and right sides of the seat is multiplied by a proportional coefficient. The conversion yields the following formula: ;in, and These are the torsional angles measured by strain sensors on the left and right sides of the seat surface, respectively, in degrees. The dimensionless proportionality coefficient is obtained through calibration of the seat's geometry and stiffness characteristics; ultimately, the real-time sitting angle is determined by... and The two components are jointly characterized.

[0074] Based on the real-time sitting angle, the muscle activation characteristics are normalized and the static tension components of the muscles are corrected to obtain the muscle tension distribution. Specifically, this includes: based on the obtained real-time sitting angle... and The obtained muscle activation characteristics were normalized, and then the static tension component of the muscles due to maintaining posture was corrected using the sitting angle to obtain the muscle tension distribution. The normalization process was performed according to the following formula: In the formula, For the first Normalized activation level of individual muscle groups, dimensionless; The root mean square (RMS) of the electromyographic amplitude of this muscle group is calculated (the aforementioned RMS is a general formula applicable to every muscle group; here...). This is to differentiate the specific values ​​for different muscle groups (this is a reasonable refinement of the notation), and the unit is microvolts; The root mean square amplitude, measured at the maximum voluntary contraction of this muscle group, is expressed in microvolts and obtained through a pre-driving static calibration process. The correction formula for the static tension component is: ;in, For the revised version The tension distribution values ​​of individual muscle groups, dimensionless; This is the global scaling factor, with a value between 0.6 and 0.8, determined through ergonomic experiments; To adjust according to real-time sitting posture angle and The table lookup result is the first The theoretical baseline values ​​of static tension for each muscle group are dimensionless and are obtained by fitting electromyographic data collected offline at multiple standard sitting angles.

[0075] After the above corrections, the muscle tension distribution reflecting fatigue-related abnormal tension is obtained. ,in It traverses the lumbar segment of the erector spinae, the upper trapezius, and the levator scapulae.

[0076] Using the road surface excitation intensity level and real-time sitting posture angle as constraints, the pressure fluctuations synchronized with road surface vibration and the pressure offsets related to sitting posture tilt are extracted from the initial body pressure characteristics to form compensated body pressure distribution change information. Specifically, in this step, a high-dimensional manifold adaptive Riesen curvature decoupling geometry algorithm is used to perform the extraction operation. First, each sampling point in the initial body pressure feature space is mapped to a five-dimensional Riemannian manifold, and the manifold coordinate vector is defined as: In the formula, For the first Pressure values ​​at each sensing point The coordinates of the pressure center are... For contact area, and These are the relative tilt angles between the waist and the backrest, and the relative rotation angles between the buttocks and the seat surface, respectively. and The dominant frequency and amplitude of the road vibration are respectively obtained; further, the Riessen curvature tensor at each point on the manifold is calculated. Its component expression is: ;in, , , , All are manifold Christofel symbols. , For manifold coordinate components, Christofel symbol coordinates The partial derivatives; This reflects the difference in the rate of change of Christofel's symbol on the manifold; The quadratic term part reflects the nonlinear interactions of the connection; when the manifold is flat (such as Euclidean space), all When a manifold bends, the non-zero components of the curvature tensor reflect the intensity and direction of the bend; when the absolute value of the Riessen curvature in a certain coordinate direction exceeds a preset threshold... When this occurs, it is determined that there is strong coupling interference in this direction; a decoupling projection operator is constructed, the formula of which is as follows: ;in, It is the identity matrix. This represents the unit feature vector corresponding to the direction where curvature exceeds the threshold. The set of all directions exceeding the threshold; the initial body pressure feature vector Projecting the vector onto a subspace orthogonal to the coupling direction yields the decoupled volume pressure feature vector. ; ultimately from Extracting the dominant frequency of road vibration Synchronous pressure fluctuation components and the angle of inclination of the sitting posture Related pressure offset components and from the original body pressure characteristics Subtracting these two components from the middle, the formula is as follows: In the formula This refers to the compensated information on changes in body pressure distribution. This information retains only the actual dynamic changes in body pressure caused by driving fatigue, while removing coupling interference from road vibrations and changes in posture. for ,in, Calculate the vector dot product operation exist Projected length in the direction, This is the initial body pressure feature vector. This represents the unit eigenvector in the direction of road vibration coupling. This refers to the finally separated road vibration interference component;

[0077] for ,in, The final separated sitting posture change interference component, This is the initial body pressure feature vector. For vector dot product operation, calculate exist Projected length in the direction, This is the unit eigenvector of the coupling direction of the sitting posture change.

[0078] It should be specifically stated that the high-dimensional manifold adaptive Riesen curvature decoupling geometry algorithm used in the above steps is only a specific implementation of this invention, used to exemplify how to separate pressure fluctuations synchronized with road vibration and pressure offsets related to sitting posture from the initial body pressure characteristics using road excitation intensity level and real-time sitting posture angle as constraints; the algorithm does not constitute any limitation on the scope of protection of the claims of this invention; the technical means described in the claims for separating pressure fluctuations synchronized with road vibration and pressure offsets related to sitting posture from the initial body pressure characteristics using road excitation intensity level and real-time sitting posture angle as constraints include, but are not limited to, independent component analysis, wavelet transform, blind source separation, adaptive filtering, principal component analysis, and the manifold geometry decoupling algorithm described in this embodiment.

[0079] In a preferred embodiment of the present invention, heart rate variability is obtained based on the heart rate signal, and a real-time fatigue index is calculated based on the heart rate variability, combined with compensated body pressure distribution change information and muscle tension distribution, including:

[0080] Extracting continuous heartbeat intervals from the heart rate signal and performing time-domain or frequency-domain analysis on these intervals to obtain heart rate variability characteristics. Specifically, this includes: extracting continuous heartbeat intervals from the heart rate signal and performing time-domain or frequency-domain analysis on these intervals to obtain heart rate variability characteristics; specifically, firstly, R-wave peak detection is performed on the heart rate signal, and the time interval between two consecutive R waves is recorded to obtain a heartbeat interval sequence, denoted as... ,in This represents the total number of heartbeat intervals in the current analysis window, expressed in times. The standard deviation of the heartbeat intervals is calculated for time-domain analysis using the following formula: ;in, This represents the standard deviation of the heartbeat interval, in milliseconds. For the first One heartbeat interval; This is the average of all heartbeat intervals within the current window, also in milliseconds.

[0081] Frequency domain analysis first resamples the intercardiac interval sequence into an equally spaced sequence using cubic spline interpolation, then calculates the power spectral density using Fast Fourier Transform, extracting the low-frequency power (0.04Hz to 0.15Hz) and high-frequency power (0.15Hz to 0.40Hz), and calculating the power ratio between the low and high frequencies. The two features mentioned above are combined into a heart rate variability feature vector. ;in, It is a two-dimensional vector. It reflects the overall degree of fluctuation in heart rate. It reflects the balance between the sympathetic and parasympathetic nervous systems.

[0082] The real-time fatigue index is obtained by fusing heart rate variability features, compensated body pressure distribution changes, and muscle tension distribution. Specifically, this includes: integrating the heart rate variability feature vector... The compensated body pressure distribution change information vector and the distribution of muscle tension By performing fusion calculations, the real-time fatigue index is obtained. The fusion calculation uses the following weighting formula: ; In the formula, It is dimensionless, and the larger its value, the higher the degree of fatigue; This represents the standard deviation of the baseline heart rate interval for a driver in an conscious state. As a benchmark for the low-frequency to high-frequency power ratio in a conscious state, The L2 norm benchmark value is used to determine the changes in body pressure distribution when the driver is in a comfortable seating position without vibration interference. These are the baseline values ​​for muscle tension in a relaxed state for the driver; all four baseline values ​​were obtained through static calibration data collected 5 minutes before driving. Representing vectors The L2 norm is calculated using the following formula: ;

[0083] in The total number of sensing points. For the first Pressure value after compensation at each sensing point; coefficient The preset weighting coefficients all range from 0.2 to 0.5 and satisfy the following conditions: The determination was made through regression analysis using a large amount of real-vehicle fatigue test data; This is a constant bias term used to calibrate the fatigue index to the normal physiological range, with a value ranging from -0.2 to 0.2. Substituting the above components into the formula yields the calculated value. This is the real-time fatigue index, which will be used by the adaptive decision-making module.

[0084] In a preferred embodiment of the present invention, the massage action sequence and seat support adjustment amount are obtained based on the real-time fatigue index, real-time sitting posture angle, and road surface excitation intensity level, including:

[0085] Based on the real-time fatigue index and real-time sitting posture angle, basic comfort requirement parameters are obtained. These parameters include massage point sequence, technique mode, base massage intensity, and base frequency. Specifically, the real-time fatigue index is divided into three intervals: 0 to 0.3 for mild fatigue, 0.3 to 0.7 for moderate fatigue, and 0.7 to 1.0 for severe fatigue. Each interval is pre-stored with a set of massage strategy parameters: the mild fatigue interval corresponds to a relaxation-oriented massage strategy, with massage points concentrated in the lumbosacral region; the moderate fatigue interval corresponds to a soothing massage strategy, with massage points covering the lumbar spine to the lower thoracic spine; and the severe fatigue interval corresponds to a stimulating massage strategy, with massage points covering the entire spine.

[0086] Furthermore, the relative tilt angle between the lumbar region and the backrest in the real-time sitting posture angle is used as a correction factor: when the tilt angle is greater than 15 degrees, it indicates that the driver is in a reclining posture, and the massage point sequence is shifted upward by 5 cm to cover the fatigue area of ​​the upper back raised due to reclining; when the tilt angle is less than 5 degrees, it indicates that the driver is in a forward-leaning posture, and the massage point sequence is shifted downward by 3 cm to concentrate on the lumbosacral region; when the tilt angle is between 5 and 15 degrees, the massage point sequence is not shifted; the method for determining the massage mode is as follows: the low-frequency to high-frequency power ratio in the heart rate variability characteristics is obtained from the fatigue quantification module. If the ratio is greater than 1.5, the fatigue type is determined to be sympathetic nerve dominant, and a soothing mode combining kneading and pressing is adopted; if the ratio is less than 0.8, it is determined to be parasympathetic nerve dominant, and a wake-up mode combining tapping and vibration is adopted; if it is between 0.8 and 1.5, a comprehensive acupressure mode is adopted.

[0087] The massage intensity base value is calculated according to a linear increment rule: the intensity base value corresponding to a real-time fatigue index of 0.3 is the initial value of 0.5 (the unit is the preset intensity level, with a maximum of 1.0). For every 0.1 increase in the fatigue index, the intensity base value increases by 0.08. When the fatigue index reaches 1.0, the upper limit of the intensity base value is 1.0. The frequency base value also increases monotonically with the fatigue index: the frequency base value is 1.2 Hz when the fatigue index is 0.3, 2.0 Hz when the fatigue index is 0.7, and 3.0 Hz when the fatigue index is 1.0. The intermediate values ​​are obtained by linear interpolation.

[0088] The basic comfort requirement parameters are fused and compensated with the road surface excitation intensity level to obtain the compensated massage action sequence. Specifically, the following steps are taken: the evaluation result of the road surface excitation intensity level is obtained from the fusion decoupling module. The level is divided into low level (1), medium level (2) and high level (3). When the road surface excitation intensity level is low, it means that the current road surface is flat and the impact of vibration on the massage effect can be ignored. Therefore, no compensation is made for the massage intensity base value and frequency base value. The output massage point sequence, technique mode, massage intensity base value and frequency base value are directly used as the final parameters of the massage action sequence. When the road surface excitation intensity level is medium, it means that there is a certain degree of bumpiness on the road surface. In order to avoid the road vibration and massage action superimposed to cause excessive shaking or discomfort to the driver's body, the massage intensity base value is multiplied by a compensation coefficient of 0.8 to reduce the actual massage intensity by 20%. At the same time, the frequency base value is multiplied by a compensation coefficient of 1.2 to increase the massage frequency by 20%. The higher frequency regular tactile stimulation is used to mask the random interference caused by road vibration.

[0089] Meanwhile, when the road surface excitation intensity level is high, it indicates severe road bumps. In this case, priority should be given to ensuring the driver's handling stability. Specific compensation measures include: multiplying the massage intensity base value by a compensation coefficient of 0.6 to reduce the massage intensity to 60% of the original level; forcibly setting the frequency base value to a fixed value of 1 Hz to avoid resonance between excessively high frequencies and road vibrations; and switching the massage mode to a gentle stroking mode, in which the massage airbags only perform slow undulating movements with extremely low pressure, no longer performing vigorous movements such as tapping or kneading. After completing the above compensation, the four elements of massage point sequence, compensated massage intensity value, compensated frequency value, and the finally determined massage mode are combined into a compensated massage action sequence for the execution control module to call.

[0090] Based on the compensated massage action sequence and the real-time sitting posture angle, the seat support adjustment amount is calculated, which includes the lumbar support protrusion amount, the side wing clamping angle, and the backrest tilt fine adjustment angle.

[0091] In a preferred embodiment of the present invention, the seat support adjustment amount is calculated based on the compensated massage action sequence and the real-time sitting posture angle. The seat support adjustment amount includes the lumbar support protrusion amount, the side wing clamping angle, and the backrest tilt fine-tuning angle, including:

[0092] The massage point sequence and technique pattern are extracted from the compensated massage action sequence. The target area for spinal support is determined based on the massage point sequence, and the required support method is determined based on the technique pattern. Specifically, the massage point sequence consists of multiple massage point coordinates distributed along the spine, with each massage point corresponding to the spatial position of an airbag or vibration unit within the seat back. The continuous interval between the minimum vertical coordinate (i.e., the uppermost massage point) and the maximum vertical coordinate (i.e., the lowermost massage point) of these massage points is marked as the target area for spinal support; this area is the vertical range where the lumbar support needs to provide focused support. The technique patterns are divided into three types: kneading mode, pressing mode, and... The system offers three modes: tapping, vibration, and gentle stroking. The kneading and pressing modes cater to drivers requiring strong and stable back support; therefore, the support method is defined as rigid support, requiring the lumbar support airbag to inflate to a high pressure to provide a definite push-up sensation. The tapping and vibration modes cater to drivers requiring dynamic stimulation rather than static support; therefore, the support method is defined as elastic cushioning, requiring the lumbar support airbag to inflate to a medium pressure while maintaining a certain degree of flexibility to absorb impact rather than counteract it. The gentle stroking mode caters to drivers requiring minimal disturbance; therefore, the support method is defined as flexible fit, requiring the lumbar support airbag to maintain only a very low base pressure.

[0093] Using the target area and support requirements as input, and combining the real-time sitting angle, the amount of lumbar support protrusion within the target area is calculated to obtain the lumbar support protrusion amount. Specifically, the lumbar support is usually composed of multiple independently controlled airbags arranged vertically, each airbag can be inflated independently to produce a local protrusion. First, the range of airbag numbers to be activated is determined according to the upper and lower boundaries of the target area: assuming there are 5 airbags arranged vertically in sequence, if the target area covers airbags 2 to 4, then airbags 2, 3, and 4 are set as active support airbags, and airbags 1 and 5 are set as transition airbags to maintain only basic pressure. Based on this, the highest point of the lumbar support protrusion is determined: this point is aligned with the center of the target area, that is, if the center of the target area is located at airbag 3, then the inflation pressure of airbag 3 is the highest, and the adjacent airbags 2 and 4 decrease by 30% successively, forming a smooth arc-shaped protrusion contour.

[0094] Further, the basic inflation pressure of the lumbar support airbag is selected based on the support method requirements: the basic inflation pressure corresponding to the rigid support requirement is 0.6 MPa, the basic inflation pressure corresponding to the elastic cushioning requirement is 0.4 MPa, and the basic inflation pressure corresponding to the flexible fit requirement is 0.2 MPa. Finally, the asymmetrical correction of the protrusion amount is made on the left and right sides based on the relative angle between the buttocks and the seat surface in the real-time sitting posture angle: when the angle of the buttocks relative to the seat surface tilting to the left is greater than 3 degrees, it is determined that the driver's pelvis is tilted to the left. At this time, the protrusion amount on the left side of the lumbar support (i.e., the left side of the backrest area) is increased by 20% on the basic value, while the protrusion amount on the right side of the lumbar support is reduced by 20% to compensate for the lateral curvature of the spine caused by the pelvic tilt; when the right side tilt is greater than 3 degrees, symmetrical treatment is performed; when the absolute value of the tilt angle does not exceed 3 degrees, the protrusion amount on the left and right sides remains the same. After the above calculations, the independent protrusion amount values ​​of each active support airbag on the left and right sides are obtained, in millimeters.

[0095] Based on the lumbar support protrusion and real-time seating angle, the clamping angle required to maintain the driver's torso lateral stability is calculated, resulting in the side wing clamping angle. Specifically, the side wing clamping mechanism is typically located on the left and right sides of the seat back, applying lateral pressure to the driver's torso through inward rotation or inflation. The side wing clamping angle is defined as the angle between the inner surface of the side wing and the vertical plane of the seat center. A larger angle indicates a stronger clamping force, with a maximum angle of 25 degrees. First, the contribution of the lumbar support protrusion to the driver's lateral positioning is evaluated: the average protrusion of all active support airbags is calculated. If the average protrusion is greater than 15 mm, it indicates that the lumbar support has provided strong lateral support, and the basic clamping angle of the side wing is set to 10 degrees. If the average protrusion is between 8 mm and 15 mm, the basic clamping angle is set to 15 degrees. If the average protrusion is less than 8 mm, the basic clamping angle is set to 20 degrees.

[0096] Further, based on the relative tilt angle between the lumbar region and the backrest, and the relative rotation angle between the buttocks and the seat surface in the real-time sitting posture angle, the comprehensive lateral offset trend of the driver's torso relative to the seat centerline is calculated: the relative rotation angle of the buttocks and the relative tilt angle of the lumbar region are weighted and averaged with weight coefficients of 0.6 and 0.4, respectively, to obtain the lateral offset index; when the index is greater than 2 degrees (shift to the left), the clamping angle of the left wing is increased by 5 degrees on the base value, while the clamping angle of the right wing is decreased by 5 degrees, so as to create a pressure difference between the left and right sides to push the driver back to the center position; when the index is less than -2 degrees (shift to the right), the opposite adjustment is made; when the absolute value of the index does not exceed 2 degrees, the clamping angles of the left and right wings are equal; finally, the clamping angles of the left and right wings are output, both of which are independent values ​​in degrees.

[0097] Based on the lumbar support protrusion, side wing clamping angle, and the real-time sitting posture angle, the backrest tilt adjustment angle is calculated. Specifically, the backrest tilt adjustment angle is a small correction amount superimposed on the driver's manually set base backrest angle. The correction range is typically -3 degrees to +3 degrees. The calculation of the adjustment angle follows three rules, and the final adjustment angle is the algebraic sum of the three rule output values. The first rule is related to the lumbar support protrusion: calculate the average protrusion of all active support airbags. When the average protrusion is greater than 20 mm, it indicates that the lumbar support is significantly pushing the driver's waist forward. To avoid the shoulders being unsupported, the backrest is slightly adjusted backward by ±2.5 degrees. When the average protrusion is between 10 mm and 20 mm, a slight adjustment of ±1.5 degrees is made. When the average protrusion is less than 10 mm, this adjustment is not performed.

[0098] The second rule relates to the side wing clamping angle: Take the larger value between the left and right side wing clamping angles. If the larger value is greater than 20 degrees, it indicates that the side wings are exerting a strong lateral wrapping on the driver's torso. To avoid chest compression caused by excessive inward movement, slightly adjust the backrest forward by -2 degrees. If the larger value is between 15 and 20 degrees, slightly adjust by -1 degree. If it is less than 15 degrees, do not make this adjustment. The third rule relates to the relative tilt angle between the lumbar region and the backrest in the real-time sitting posture angle: Compare this tilt angle with the preset comfort threshold range (12 to 18 degrees). If the tilt angle is greater than 18 degrees, it indicates that the driver is overly... If the driver leans back, the backrest is slightly adjusted forward by -1.5 degrees. If the tilt angle is less than 12 degrees, it indicates that the driver is leaning too far forward, so the backrest is slightly adjusted backward by +1.5 degrees. If the tilt angle is between 12 and 18 degrees, this adjustment is not performed. The adjustment angles obtained from the above three rules are algebraically summed to obtain the final backrest tilt adjustment angle. This adjustment angle is added to the basic backrest angle manually set by the driver to form an execution command sent to the backrest adjustment mechanism. If the summed angle exceeds the mechanical travel range of the actuator, which is usually between -5 and +30 degrees, a limiting process is performed, and the closest boundary value is taken.

[0099] In a preferred embodiment of the present invention, the step of controlling the massage unit in the seat to perform massage according to the massage action sequence, and simultaneously driving the lumbar support adjustment mechanism, side wing clamping mechanism and backrest adjustment mechanism of the car massage seat to perform adjustment according to the seat support adjustment amount, includes:

[0100] According to the massage point sequence, massage intensity value, frequency value, and technique mode in the compensated massage action sequence, the massage unit in the seat is driven to perform massage along the driver's spine, forming a massage execution feedback signal. Specifically, the massage unit in the seat consists of multiple airbags or vibration motors arranged along the spine, with each massage point corresponding to an independent actuator. The massage point sequence is first analyzed into the activation time and order of each massage point. The corresponding actuators are triggered sequentially according to the path specified in the sequence (e.g., from the lumbosacral region upwards to the shoulder and neck). The massage intensity value of each massage point is achieved by controlling a pneumatic proportional valve or vibration amplitude adjustment circuit: the greater the intensity value, the higher the airbag inflation pressure or the greater the vibration motor duty cycle. The frequency value determines the rhythm of the massage action. For airbag massage units, the frequency value controls the speed of inflation and deflation switching. For vibration massage units, the frequency value directly controls the driving frequency of the vibration motor. The technique mode specifies the action waveform of each massage point during activation. The kneading mode corresponds to the pressure change of a sine wave, the tapping mode corresponds to the rapid impact of a pulse waveform, and the stroking mode corresponds to a trapezoidal waveform that rises and then falls slowly.

[0101] During execution, pressure sensors or acceleration sensors embedded in the massage unit collect the actual force or vibration amplitude of each massage point in real time. These measured values ​​are compared with the preset force and frequency values ​​of the massage point to form a massage execution feedback signal that includes the actual execution deviation and the progress of the action at each massage point. This signal is continuously updated at a frequency of once every 50 milliseconds.

[0102] Based on the massage execution feedback signal and the lumbar support protrusion amount, the lumbar support adjustment mechanism is driven to perform lumbar support protrusion adjustment to obtain the lumbar support adjustment completion state. Specifically, this includes: based on the formed massage execution feedback signal and the lumbar support protrusion amount calculated by the adaptive decision module, the lumbar support adjustment mechanism is driven to perform lumbar support protrusion adjustment to obtain the lumbar support adjustment completion state; the lumbar support adjustment mechanism consists of multiple independent airbags arranged vertically, each airbag can be independently inflated and deflated by an air pump and a solenoid valve. First, the vertical position of the current massage point is extracted from the massage execution feedback signal: if the current massage point is located in the lumbosacral region, corresponding to the 1st to 2nd airbag, then it is determined that the lumbar support should first support this region; if the massage point is located in the lower thoracic spine, corresponding to the 3rd to 4th airbag, then the target area supported by the lumbar support is moved upward synchronously.

[0103] Further, based on the lumbar support protrusion (including the target protrusion height of each airbag, in millimeters), a target inflation pressure value for each airbag is generated. Closed-loop control is used to make the current pressure of each airbag approach the target pressure. While performing lumbar support adjustment, the massage execution feedback signal is continuously monitored: if the actual force value of a certain massage point is detected to be lower than the preset value for more than 0.5 seconds, it is determined that the lumbar support is insufficient, resulting in the massage unit not fitting tightly against the driver's back. At this time, the protrusion of the corresponding lumbar support airbag is automatically increased by 10% as compensation. When the current pressure of all lumbar support airbags reaches the ±5% error range of the target pressure value and remains within the error range for 0.3 seconds, the lumbar support adjustment is marked as true, and the actual final protrusion of each airbag is recorded as part of the lumbar support adjustment completion status.

[0104] Based on the lumbar support adjustment completion status and the side wing clamping angle, the side wing clamping mechanism is driven to perform side wing clamping to obtain the side wing adjustment completion status. Specifically, the side wing clamping mechanism is located on the left and right sides of the seat back, and each side wing has a built-in rotary motor or airbag, which can independently control the inward rotation angle or inflation amount of the side wing. Before the lumbar support adjustment completion status is true, the side wing clamping mechanism maintains a zero position (minimum angle) to avoid improper compression of the driver by the side wings when the lumbar support is not in place. Once the lumbar support adjustment completion sign is received, the left wing clamping angle and the right wing clamping angle output by the adaptive decision module are immediately read. These two angles are usually not equal to correct the driver's lateral deviation trend.

[0105] The driving process adopts a step-by-step approach: first, both side wings are rotated simultaneously from zero to the target angle at a speed of 5 degrees per second, and the side wing angle sensor value is read every 0.2 seconds; when the angle of a side wing reaches within ±1 degree of the target value, the drive motor of that side is stopped and the position is maintained; during the side wing clamping process, the change in the driver's body pressure distribution is monitored in real time: if a sudden and significant increase in body pressure on one side is detected, exceeding 50% of the average pressure on that side, it is determined that the side wing is over-clamped, and the side wing on that side is immediately moved back 2 degrees and slowly pushed forward again. When the angles of both side wings are stable within ±1 degree of the target value for more than 0.5 seconds, the side wing adjustment is marked as true, and the actual clamping angles on both sides are recorded.

[0106] Based on the side wing adjustment completion status and the backrest pitch fine-tuning angle, the backrest adjustment mechanism is driven to perform backrest pitch fine-tuning. Specifically, this includes: based on the obtained side wing adjustment completion status and the backrest pitch fine-tuning angle, the backrest adjustment mechanism is driven to perform backrest pitch fine-tuning to obtain the backrest pitch fine-tuning completion status. Specifically, the backrest adjustment mechanism is usually driven by an electric lead screw or an angle adjuster motor, which can realize continuous adjustment of the backrest pitch angle relative to the seat surface. Before the side wing adjustment completion status is true, the backrest maintains the basic angle manually set by the driver and does not perform fine-tuning actions. Once the side wing adjustment completion flag is received, the backrest pitch fine-tuning angle (in degrees, a positive value indicates tilting backward and a negative value indicates tilting forward) output by the adaptive decision module is read, and this fine-tuning angle is added to the current basic backrest angle to obtain the target backrest angle.

[0107] The backrest adjustment mechanism moves towards the target angle at an angular velocity of 1.5 degrees per second, reading the backrest angle sensor value every 0.1 seconds during the movement. When the absolute value of the difference between the current backrest angle and the target angle is less than 0.5 degrees, the drive motor stops and the angle adjuster is locked. During the backrest fine-tuning process, the massage execution feedback signal and the lumbar support pressure signal are monitored simultaneously. If a decrease of more than 20% in the massage intensity value is detected, it indicates that the change in the backrest angle may cause a change in the fit between the driver and the seat. At this time, the backrest movement is paused for 0.5 seconds to allow the driver to readjust their sitting posture before continuing the movement. When the backrest angle stabilizes within ±0.5 degrees of the target angle for more than 0.5 seconds, the backrest pitch fine-tuning is marked as complete, and the final backrest angle is fed back to the vehicle comfort control terminal. At this point, the massage execution, lumbar support adjustment, side wing clamping, and backrest fine-tuning are all completed, and the adaptive adjustment system enters the steady-state comfort support mode.

[0108] Embodiments of the present invention also provide a computing device, including: a processor and a memory storing a computer program, wherein the computer program, when executed by the processor, performs the functions of the system described above. All implementations in the above system embodiments are applicable to this embodiment and can achieve the same technical effects.

[0109] Embodiments of the present invention also provide a computer-readable storage medium storing instructions that, when executed on a computer, cause the computer to perform the functions of the system described above. All implementations in the above system embodiments are applicable to this embodiment and can achieve the same technical effects.

[0110] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. An adaptive adjustment system for a car massage seat, characterized in that, include: The multi-source sensor acquisition module is used to acquire in real time the body pressure distribution signal in the seat surface, the road vibration signal transmitted by the vehicle chassis, the deformation signal of the seat frame, the driver's muscle tension signal, and the heart rate signal. The fusion decoupling module is used to jointly process the body pressure distribution signal, road vibration signal, deformation signal and muscle tension signal, separate the body pressure coupling interference caused by road unevenness and sitting posture changes, and obtain the driver's real-time sitting posture angle, road excitation intensity level and muscle tension distribution to form compensated body pressure distribution change information. The fatigue quantification module is used to obtain heart rate variability based on the heart rate signal, and calculate the real-time fatigue index based on the heart rate variability, combined with the compensated body pressure distribution change information and the muscle tension distribution. An adaptive decision-making module is used to obtain the massage action sequence and seat support adjustment amount based on the real-time fatigue index, real-time sitting posture angle and road surface excitation intensity level. The massage sequence includes a sequence of massage points that dynamically slide along the driver's spine, massage intensity, frequency, and technique; the seat support adjustment includes lumbar support protrusion, side wing clamping angle, and backrest tilt adjustment angle. The execution control module is used to control the massage unit in the seat to perform massage according to the massage action sequence, and simultaneously drive the lumbar support adjustment mechanism, side wing clamping mechanism and backrest adjustment mechanism from the car massage seat to perform adjustment according to the seat support adjustment amount.

2. The adaptive adjustment system for a car massage seat according to claim 1, characterized in that, Real-time acquisition of body pressure distribution signals within the seat surface, road vibration signals transmitted from the vehicle chassis, seat frame deformation signals, driver muscle tension signals, and heart rate signals, including: By collecting the pressure distribution caused by body contact through multiple preset sensing points on the seat surface corresponding to the driver's lumbar spine, thoracic spine, sacrum and ischial tuberosity, a body pressure distribution signal is formed. Mechanical vibrations caused by uneven road surfaces are picked up from the connection point between the seat frame and the vehicle chassis. The mechanical vibrations are converted into acceleration signals. The acceleration signals are then subjected to bandpass filtering and root mean square calculation to obtain the road vibration signal. Detect minute deformations of the seat frame under the driver's weight and dynamic loads, and generate deformation signals; Electromyographic activity information of target muscle groups in the driver's lower back and neck is obtained, and the electromyographic activity information is processed by full-wave rectification and moving average filtering to obtain muscle tension signal; Heart rate-related physiological electrical signals or photoplethysmography signals are collected from the driver's body parts. The heart rate signal is obtained by detecting and extracting the heartbeat interval from the physiological electrical signals or the pulse interval from the photoplethysmography signals and converting them.

3. The adaptive adjustment system for a car massage seat according to claim 2, characterized in that, The body pressure distribution signal, road vibration signal, deformation signal, and muscle tension signal are jointly processed to separate the body pressure coupling interference caused by road unevenness and changes in sitting posture, obtaining the driver's real-time sitting posture angle, road excitation intensity level, and muscle tension distribution, forming compensated body pressure distribution change information, including: The pressure center coordinates, contact area, and pressure values ​​at each preset sensing point are extracted from the body pressure distribution signal to form initial body pressure characteristics; the bending angle and torsion of the seat frame are extracted from the deformation signal to obtain seat deformation characteristics; the road vibration signal is subjected to spectral analysis to extract the dominant frequency and amplitude to obtain vibration characteristics; the root mean square of electromyographic amplitude is extracted from the muscle tension signal to obtain muscle group activation characteristics. The vibration characteristics are correlated with the seat deformation characteristics, including separating the seat frame response components caused by road excitation, and evaluating the road excitation intensity level based on the dominant frequency and amplitude. Based on the road excitation intensity level and seat deformation characteristics, the relative tilt angle between the driver's waist and the backrest and the relative rotation angle between the buttocks and the seat surface are calculated to obtain the real-time sitting posture angle. Based on the real-time sitting angle, the muscle activation characteristics are normalized and the static tension component of the muscle group is corrected to obtain the muscle tension distribution. Using the road surface excitation intensity level and real-time sitting posture angle as constraints, the pressure fluctuations synchronized with road surface vibration and the pressure offsets related to sitting posture tilt are extracted from the initial body pressure characteristics to form compensated body pressure distribution change information.

4. The adaptive adjustment system for a car massage seat according to claim 3, characterized in that, Based on the heart rate signal, heart rate variability is obtained. Based on this heart rate variability, combined with compensated changes in body pressure distribution and muscle tension distribution, a real-time fatigue index is calculated, including: Extract continuous heartbeat intervals from the heart rate signal, and perform time-domain or frequency-domain analysis on the continuous heartbeat intervals to obtain heart rate variability characteristics. The real-time fatigue index is obtained by fusing the heart rate variability characteristics, the compensated body pressure distribution change information, and the muscle tension distribution.

5. The adaptive adjustment system for a car massage seat according to claim 4, characterized in that, Based on the real-time fatigue index, real-time sitting posture angle, and road surface excitation intensity level, the massage action sequence and seat support adjustment amount are obtained, including: Based on the real-time fatigue index and real-time sitting angle, basic comfort requirement parameters are obtained; among which, basic comfort requirement parameters include massage point sequence, massage technique mode, massage intensity base value and frequency base value; The basic comfort requirement parameters are fused and compensated with the road surface excitation intensity level to obtain the compensated massage action sequence; Based on the compensated massage action sequence and the real-time sitting posture angle, the seat support adjustment amount is calculated, which includes the lumbar support protrusion amount, the side wing clamping angle, and the backrest tilt fine adjustment angle.

6. The adaptive adjustment system for a car massage seat according to claim 5, characterized in that, Based on the compensated massage motion sequence and the real-time sitting posture angle, the seat support adjustment amount is calculated. This seat support adjustment amount includes the lumbar support protrusion amount, the side wing clamping angle, and the backrest tilt fine-tuning angle, including: The massage point sequence and technique pattern are extracted from the compensated massage action sequence. The target area for spinal support is determined based on the massage point sequence, and the support method requirement is determined based on the technique pattern. Using the target area and support method requirements as input, and combining the real-time sitting angle, the amount of lumbar support protrusion within the target area is calculated to obtain the lumbar support protrusion. Based on the lumbar support protrusion and the real-time sitting angle, the clamping angle required to maintain the lateral stability of the driver's torso is calculated, and the side wing clamping angle is obtained. Based on the lumbar support protrusion, the side wing clamping angle, and the real-time sitting posture angle, the backrest tilt adjustment angle is calculated to obtain the backrest tilt adjustment angle.

7. The adaptive adjustment system for a car massage seat according to claim 6, characterized in that, The process of controlling the massage units within the seat to perform massage according to the massage action sequence, and simultaneously driving the lumbar support adjustment mechanism, side wing clamping mechanism, and backrest adjustment mechanism of the car massage seat to perform adjustments according to the seat support adjustment amount, includes: According to the massage point sequence, massage intensity value, frequency value and technique mode in the compensated massage action sequence, the massage unit in the seat is driven to perform massage along the driver's spine, forming a massage execution feedback signal; Based on the massage execution feedback signal and the lumbar support protrusion amount, the lumbar support adjustment mechanism is driven to perform lumbar support protrusion adjustment to obtain the lumbar support adjustment completed state. Based on the lumbar support adjustment completion state and the side wing clamping angle, drive the side wing clamping mechanism to perform side wing clamping, and obtain the side wing adjustment completion state; Based on the completion status of the side wing adjustment and the backrest pitch fine-tuning angle, the backrest adjustment mechanism is driven to perform backrest pitch fine-tuning.

8. A computing device, characterized in that, include: One or more processors; A storage device for storing one or more programs that, when executed by one or more processors, cause the one or more processors to perform the functions of the system as described in any one of claims 1 to 7.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a program that, when executed by a processor, performs the functions of the system as described in any one of claims 1 to 7.

Images