Driving performance parameter adjusting method and system

Abstract

The invention relates to a driving performance parameter adjusting method and system, and belongs to the technical field of vehicle control. The method is applied to a vehicle and comprises the steps that electromyographic signals of all passengers in the vehicle are collected; generating a vehicle comfort score based on the electromyographic signals of all passengers; and according to the vehicle comfort level score, the driving performance parameters of the vehicle are adjusted. The electromyographic signals of all passengers serve as an objective input source of comfort, the muscle tension state of the human body caused by jolting, pause or power impact in the vehicle running process is directly reflected, and therefore the actual comfort feeling of the passengers is truly captured in time. On this basis, the system actively adjusts driving performance parameters based on the generated vehicle comfort score to improve ride comfort. According to the mechanism, the problems of driving performance parameter adjustment lag, amplitude failure or direction error and the like caused by lack of human body feedback in a traditional method are effectively solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of vehicle control, in particular to a driving performance parameter adjustment method and system. BACKGROUND

[0002] With the rapid development of intelligent and electric technologies of automobiles, modern vehicles are generally equipped with various adjustable driving performance parameters, such as driving modes, suspension characteristics (including height and damping) and energy recovery intensity, etc. At present, the development trend of the industry is to enable vehicle control systems to automatically adjust the above parameters based on the current driving scene, so as to ensure the comfort experience of passengers in various road conditions.

[0003] To this end, some vehicle models introduce variable damping suspensions and road preview systems, which adjust the suspension state in advance by sensing the road conditions ahead, so as to suppress body vibration and improve ride comfort. However, such solutions mainly rely on road condition information for decision-making, and are difficult to effectively cope with the reduction of comfort caused by aggressive driving style or low-grade road surface and other complex factors.

[0004] To make up for this deficiency, some existing vehicle models begin to use vehicle dynamic indicators such as vehicle speed change rate and steering angle fluctuation to indirectly infer ride comfort, and accordingly trigger parameter adjustment. However, there is a large deviation between such indirect evaluation method and the real comfort experience of passengers, resulting in a lag in the timing and an improper range of adjustment of driving performance parameters, and even in some working conditions, it produces reverse interference, resulting in poor passenger experience. SUMMARY

[0005] The present application provides a driving performance parameter adjustment method and system, which aims to monitor the electromyographic signals through wearable devices, analyze the comfort of passengers, and comprehensively adjust the driving performance parameters of vehicles, so as to overcome the lag and improper range problems of existing technical methods in adjusting the driving performance parameters of vehicles, and thus improve the riding experience of passengers.

[0006] To achieve the above-mentioned purpose, the present application adopts the following technical solutions: The present application provides a driving performance parameter adjustment method applied to a vehicle, the method comprising: Collecting electromyographic signals of all passengers in the vehicle; Generating a vehicle comfort score based on the electromyographic signals of all passengers; Adjusting the driving performance parameters of the vehicle according to the vehicle comfort score.

[0007] In the above embodiments, the application realizes a complete technical closed loop of direct perception, quantitative evaluation and active optimization of ride comfort by combining real-time collection of occupant surface electromyography signals with dynamic adjustment of vehicle driving performance parameters. Compared with the prior art which relies on indirect indicators such as vehicle body vibration and steering angle change to infer smoothness, the application takes the electromyography signals of all occupants as the objective input source of comfort, directly reflecting the muscle tension state of the human body caused by bumps, jerks or power impact during vehicle operation, thereby truly and timely capturing the actual comfort perception of the occupants. On this basis, the system actively adjusts the driving performance parameters based on the generated vehicle comfort score to improve ride smoothness. This mechanism effectively overcomes the problems of adjustment lag, improper amplitude or wrong direction caused by the lack of human feedback in traditional methods, significantly improving the adaptive comfort protection capability of the vehicle under complex road conditions and diverse driving styles.

[0008] In addition, by fusing the electromyography signals of all occupants in the vehicle, the application avoids the limitations of focusing only on the driver or a single occupant, ensures that the control decision is based on the comfort needs of the most sensitive occupant, and truly realizes a people-oriented intelligent comfort experience. In some embodiments, the method of generating a vehicle comfort score includes: Generating different individual comfort scores based on the electromyography signals of each occupant; Fusing the individual comfort scores of all occupants to obtain the vehicle comfort score.

[0009] In the above embodiments, the application realizes comprehensive representation of the comfort state of all occupants in the vehicle by generating individual comfort scores for each occupant and fusing them, avoiding control bias caused by relying only on the signals of a single occupant (such as the driver), making the vehicle control decision more global and fair. In some embodiments, the method of generating an individual comfort score includes: Preprocessing the electromyography signals to obtain effective physiological signals; Based on the effective physiological signals, obtaining time sequence features representing muscle state; Based on the time sequence features, generating a corresponding individual comfort score.

[0010] In the above embodiments, the application first preprocesses the electromyography signals to filter out noise and interference, obtaining high-quality effective physiological signals, and further extracts time sequence features that can represent the muscle state of the occupants, so that the data input to the multi-classification model has clear physiological significance and time dynamic characteristics; this method effectively improves the consistency between the individual comfort score and the real subjective perception of the occupants, avoiding misjudgment caused by directly using the original electromyography signals, thereby providing a reliable basis for the accurate adjustment of vehicle driving performance parameters.

[0011] In some embodiments, the effective physiological signal is input into a time sequence neural network to obtain the time sequence feature; or, a feature value is calculated based on the effective physiological signal, and the feature value is taken as the time sequence feature. The time sequence feature is input into a multi-classification model to obtain the individual comfort level score.

[0012] In the above embodiments, the present application significantly improves the applicability and deployment flexibility of the comfort level evaluation model by providing two parallel time sequence feature acquisition paths. On the one hand, the time sequence neural network can automatically learn high-order dynamic representations from the original electromyography signal, which is suitable for vehicle-mounted computing platforms with sufficient computing power, and realizes end-to-end high-precision comfort level prediction; on the other hand, the calculation method based on the root mean square value, integral electromyography value or modified average absolute value has low complexity, high interpretability and strong real-time performance, which is convenient for efficient operation in resource-constrained embedded systems. The two paths share the subsequent multi-classification model structure, which not only guarantees the uniformity of the system architecture, but also allows flexible selection of feature extraction strategies according to different vehicle hardware configurations, balancing cost and efficiency without sacrificing evaluation performance, thereby providing reliable, robust and practical physiological feedback for adaptive adjustment of vehicle driving performance parameters. In some embodiments, the method of fusion processing is to calculate the mean or minimum value of the individual comfort level scores of all passengers, and take the calculation result as the vehicle comfort level score.

[0013] In the above embodiments, the present application provides two optional schemes by calculating the mean or minimum value of the individual comfort level scores of all passengers as the vehicle comfort level score. The mean value can reflect the average comfort level of all passengers in the vehicle, which is suitable for conventional scenarios that pursue overall driving and riding quality; the minimum value takes the feeling of the least comfortable passenger as the decision basis, effectively avoiding strong discomfort of individual passengers due to road impact or vehicle dynamic response, which is particularly suitable for family travel or scenarios that focus on protecting vulnerable passengers (such as the elderly and children). In addition, this method is simple to calculate and has low resource overhead, which is convenient for real-time implementation in vehicle controllers, while considering system fairness and humanistic care, significantly improving the adaptability of intelligent chassis adjustment and user satisfaction. In some embodiments, the driving performance parameters include suspension damping, energy recovery intensity and driving mode; the method of adjusting the driving performance parameters of the vehicle includes: Mapping the vehicle comfort level score to a preset comfort level level interval; According to the comfort level level interval, setting the gear position of the suspension damping, the energy recovery intensity and / or the driving mode.

[0014] In the above embodiments, the vehicle comfort score is mapped to a preset comfort level interval, and a corresponding parameter adjustment strategy is adopted according to the interval, so that the adjustment of the driving performance parameters can dynamically adapt to the change gradient of the passenger comfort state, avoiding the adjustment lag or excessive response caused by a single fixed rule. The grading mapping mechanism provides a structured decision-making framework for the system, which supports strong intervention measures in low comfort and allows conservative or maintenance strategies in high comfort, thereby ensuring the riding experience while improving the flexibility and robustness of the control system. In some embodiments, the comfort level interval includes a first interval, a second interval, a third interval, and a fourth interval. When the vehicle comfort score falls into the first interval, the suspension damping, the energy recovery intensity, and the gear of the driving mode are all set to their respective preset high-comfort gears. When the vehicle comfort score falls into the second interval, the gear change amounts of the suspension damping, the energy recovery intensity, and the driving mode are determined according to a first adjustment strategy, and the gears of the corresponding parameters are adjusted based on the determined gear change amounts. When the vehicle comfort score falls into the third interval, the gear change amounts of the suspension damping, the energy recovery intensity, and the driving mode are determined according to a second adjustment strategy, and the gears of the corresponding parameters are adjusted based on the determined gear change amounts. When the vehicle comfort score falls into the fourth interval, the suspension damping, the energy recovery intensity, and the gear of the driving mode remain unchanged.

[0015] In the above embodiments, the vehicle comfort score is divided into four preset level intervals, and different adjustment strategies are adopted for different intervals to achieve graded control of driving performance parameters. In the second and third intervals, the system no longer uses the traditional whole switching to the preset mode, but uses the independent gear change amount of each parameter as the decision variable, calculates the gear number that the suspension damping, energy recovery intensity, and driving mode should move towards high comfort, and adjusts the gears accordingly. This avoids excessive or insufficient adjustment caused by forced synchronous switching of all parameters. At the same time, by setting the highest comfort gear in the first interval and prohibiting parameter disturbance in the fourth interval, the basic riding experience in extreme discomfort scenarios and the system stability in high comfort state are effectively guaranteed, improving the adaptability and user perception consistency of the vehicle comfort control. In some embodiments, the first adjustment strategy includes a first target and a first condition. The first target is that a weighted comfort degree improvement benefit obtained by adding the gear change amount of the suspension damping multiplied by two, the gear change amount of the energy recovery strength multiplied by zero point five, and the gear change amount of the driving mode, is subtracted by a performance loss penalty calculated by multiplying the first coefficient by the sum of the gear change amount of the suspension damping, the gear change amount of the energy recovery strength, and the gear change amount of the driving mode, and the final result reaches a maximum value; The first condition is that only the suspension damping, the energy recovery strength, and the driving mode are allowed to be adjusted in the direction of the high comfort gear, and at most only two of the suspension damping, the energy recovery strength, and the driving mode are adjusted; When the first target and the first condition are both met, the gear change amount of the suspension damping, the gear change amount of the energy recovery strength, and the gear change amount of the driving mode are adjusted according to the calculation result; When the first target and the first condition cannot be met at the same time, the gears of the suspension damping, the energy recovery strength, and the driving mode are all set to the high comfort gear.

[0016] In the above embodiment, the first adjustment strategy of maximizing the weighted comfort degree benefit is adopted in the second interval, and at most only two driving performance parameters are adjusted in the same adjustment period, so that the system can cooperatively optimize the suspension damping, the energy recovery strength, and the driving mode in a controlled manner when the passenger comfort is obviously insufficient but not extremely poor, and the balance between comfort and controllability is achieved. The suspension damping is given the highest weight because it has the greatest impact on comfort, the energy recovery strength is given the second highest weight, and the driving mode is given the lowest weight, so as to prioritize the chassis vibration filtering performance. In addition, when there is no feasible solution to the optimization problem, the system automatically reverts to the preset comfort configuration, ensuring the robustness and safety of the control strategy. In some embodiments, the second adjustment strategy includes a second target and a second condition; The second target is that a weighted comfort degree improvement benefit obtained by adding the gear change amount of the suspension damping multiplied by two, the gear change amount of the energy recovery strength multiplied by zero point five, and the gear change amount of the driving mode, is subtracted by a performance loss penalty calculated by multiplying the first coefficient by the sum of the gear change amount of the suspension damping, the gear change amount of the energy recovery strength, and the gear change amount of the driving mode, and the final result reaches a maximum value; The second condition is that only the suspension damping, the energy recovery strength, and the driving mode are allowed to be adjusted in the direction of the high comfort gear, and at most only one of the suspension damping, the energy recovery strength, and the driving mode is adjusted; When the second target and the second condition are simultaneously established, then the calculated gear change amount of the suspension damping, the gear change amount of the energy recovery intensity and the gear change amount of the driving mode are adjusted; When the second target and the second condition cannot be simultaneously established, the gears of the suspension damping, the energy recovery intensity and the driving mode are all set to the high comfort gear.

[0017] In the above embodiment, the application adopts the same target function as the second interval but more stringent parameter adjustment limit in the third interval, that is, only one driving performance parameter is allowed to be adjusted in the same control cycle, so that the system implements the minimum necessary intervention in the working condition where the passenger comfort is basically acceptable: the single parameter with the highest marginal benefit for improving current comfort (such as suspension damping) is preferentially selected for fine tuning, while the remaining parameters remain unchanged, thereby avoiding the sudden change of driving feeling or energy efficiency fluctuation caused by the simultaneous change of multiple parameters; this strategy further optimizes the riding experience while maintaining the original driving style of the user and the dynamic characteristics of the vehicle; in addition, when there is no feasible solution to the optimization problem, it also falls back to the preset comfort configuration, ensuring the control robustness in all working conditions. In addition, the application also provides a driving performance parameter adjustment system applied to a vehicle, comprising: An electromyographic signal acquisition module, which is configured to acquire electromyographic signals of all passengers in the vehicle; A comfort evaluation module, which is configured to generate a vehicle comfort score based on the electromyographic signals of all passengers; A parameter adjustment execution module, which is connected with the comfort evaluation module; the parameter adjustment execution module is configured to adjust the driving performance parameters of the vehicle according to the vehicle comfort score.

[0018] In the above embodiment, the application constructs a closed-loop control system composed of an electromyographic signal acquisition module, a signal preprocessing module, a comfort evaluation module and a parameter adjustment execution module, realizing end-to-end control from passenger physiological state perception to adaptive adjustment of driving performance parameters: the electromyographic signals of all passengers in the vehicle are used as the objective physiological basis for comfort evaluation, overcoming the limitations of traditional vehicle kinematics indicators such as vehicle speed change rate and steering angle fluctuation, which cannot truly reflect the muscle tension state of the human body; the effective physiological signals obtained after filtering noise and interference by the signal preprocessing module provide high signal-to-noise ratio input for the comfort evaluation module, significantly improving the accuracy of the comfort score; the parameter adjustment execution module dynamically adjusts the suspension damping, energy recovery intensity and driving mode based on the score, so that the vehicle can respond to the overall comfort needs of the passenger group in real time, actively optimizing the riding experience without human intervention, and effectively improving the comfort of the vehicle.

[0019] It should be understood that the foregoing general description and the following detailed description are only exemplary and explanatory and are not restrictive of the application. BRIEF DESCRIPTION OF DRAWINGS

[0020] Figure 1 is a flowchart of a driving performance parameter adjustment method provided by an embodiment of the application; Figure 2 is a flowchart of a method for generating a vehicle comfort score provided by an embodiment of the application; Figure 3 is a flowchart of a method for generating an individual comfort score provided by an embodiment of the application; Figure 4 is a flowchart of a method for adjusting a vehicle driving performance parameter provided by an embodiment of the application; Figure 5 is an architectural schematic diagram of a driving performance parameter adjustment system provided by an embodiment of the application; Figure 6 is a work flow diagram of a driving performance parameter adjustment system provided by an embodiment of the application; Figure 7 is an architectural schematic diagram of a feature extraction sub-model provided by an embodiment of the application.

[0021] In the above figures: 100, a car machine system; 110, a signal preprocessing module; 120, a signal framing module; 130, a comfort evaluation module; 140, a parameter adjustment execution module; 200, a wearable device; 210, a wireless transmission device. DETAILED DESCRIPTION

[0022] In the description of the present application, it should be understood that the terms "center", "longitudinal", "transverse", "length", "width", "thickness", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", "clockwise", "counterclockwise", "axial", "radial", "circumferential" and the like indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, and are only for the convenience of describing the present application and simplifying the description, and therefore cannot be understood as indicating or implying that the devices or elements referred to must have a particular orientation, be constructed and operated in a particular orientation, and therefore cannot be understood as limiting the present application. In the present application, unless specifically defined otherwise, the terms "mount", "connected", "connecting", "fixed", and "unfixed" shall be construed broadly, for example, can be fixed connection, can be detachable connection, or integrated; can be mechanical connection, or electrical connection or communication with each other; can be directly connected, or indirectly connected through an intermediate medium, or the internal communication of two elements or the interaction between two elements, unless otherwise specifically defined. For those skilled in the art, the specific meaning of the above terms in the present application can be understood according to the specific circumstances.

[0023] In the present application, unless specifically defined otherwise, the first feature is "on" or "under" the second feature can be that the first and second features are in direct contact, or the first and second features are in indirect contact through an intermediate medium. Moreover, the first feature "above", "over" and "on" the second feature can be that the first feature is directly above or obliquely above the second feature, or only indicates that the first feature is higher than the second feature in horizontal height. The first feature "below", "under" and "under" the second feature can be that the first feature is directly below or obliquely below the second feature, or only indicates that the first feature is lower than the second feature in horizontal height.

[0024] In the present application, the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" means that the specific features, structures, materials or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present application. In the present specification, the illustrative description of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any appropriate manner in any one or more embodiments or examples. In addition, those skilled in the art can combine and combine different embodiments or examples described in the specification and the features of different embodiments or examples without contradiction.

[0025] In addition, if "and / or" appears throughout the text, it means that it includes three parallel schemes, for example, "A and / or B" includes A scheme, or B scheme, or A and B scheme.

[0026] In the following, the present application is specifically described by exemplary embodiments. However, it should be understood that the elements, structures and features in one embodiment can also be beneficially combined into other embodiments without further description.

[0027] With the rapid development of intelligent and electric technologies, modern vehicles are generally equipped with a variety of adjustable driving performance parameters, such as driving modes, suspension characteristics (including height and damping), and energy recovery intensity (including brake energy recovery and sliding energy recovery). Currently, the industry trend is to enable vehicle control systems to automatically adjust these parameters based on the current driving scenario, thereby ensuring the comfort experience of passengers in various road conditions.

[0028] To this end, some models have introduced variable damping suspensions and road preview systems, which use cameras, radars, or high-precision maps to sense the road conditions ahead and actively adjust the suspension damping or height before the vehicle enters a bumpy area, to suppress body vibration and improve ride comfort. However, such solutions are highly dependent on the accuracy and completeness of external environmental information, and their adjustment logic is essentially based on prediction rather than actual ride feedback. Once encountering unrecognised sudden obstacles (such as potholes, speed bumps), continuous slight bumps, or unstructured roads (such as rural dirt roads), the system often fails to respond in time, resulting in a lag in adjustment action relative to the actual discomfort. More critically, such methods completely ignore the direct impact of driver operating style (such as frequent hard acceleration and heavy braking) on passenger comfort, and these behaviors can cause significant body movement and discomfort even on good roads.

[0029] To compensate for the shortcomings of pure road condition driving strategies, some models instead use vehicle dynamic indicators such as vehicle speed change rate, longitudinal / lateral acceleration, and steering angle fluctuation to indirectly infer ride comfort, and trigger parameter adjustments accordingly. However, these indicators only reflect the motion state of the vehicle itself, and there is no stable and universal correspondence between them and the subjective comfort of passengers. For example, when driving at a constant speed through a continuous small bump road, the vehicle acceleration signal may change slightly, but the passengers may still feel significant discomfort due to high-frequency resonance. Due to the lack of direct criteria for the actual comfort state of passengers, the control system has difficulty accurately determining the necessity, timing, and degree of adjustment, often resulting in a serious lag in adjustment action, i.e., the response is initiated only after the passengers have been in an uncomfortable state for a long time; or the adjustment amplitude is inappropriate; or even in multi-factor coupled working conditions, the adjustment direction deviates, optimizing only one dimension while exacerbating discomfort in other dimensions. Ultimately, it is difficult to achieve precise, timely, and appropriate comfort protection in complex and variable actual driving environments, and it is difficult to ensure that passengers are always in an ideal comfort interval.

[0030] Based on this, the present application provides a driving performance parameter adjustment method and system, which aims to monitor the electromyographic signal through a wearable device 200, analyze the comfort of passengers, and comprehensively adjust the driving performance parameters of the vehicle, to overcome the lag and inappropriate amplitude problems in adjusting the driving performance parameters of the vehicle in the prior art method, and thereby improve the ride experience of passengers.

[0031] Hereinafter, the embodiments of the present application will be described in detail with reference to the accompanying drawings.

[0032] As shown in the accompanying Figures 1 to 4 In one illustrative embodiment of the present application, a ride performance parameter adjustment method is disclosed, which is applied to a vehicle, and the method comprises the following steps: S1, collecting the surface electromyography signals of all passengers in the vehicle.

[0033] Further, the data collected by the present application is the surface electromyography signals of all passengers.

[0034] In some embodiments, the present application monitors the surface electromyography signals of passengers in the vehicle in real time through the wearable device 200, and analyzes the comfort state of the passengers based on the signals. Specifically, the passengers wear wearable devices with electromyography collection function, such as electromyography bracelets or smart armbands, which are attached to specific muscle groups of the human body, such as forearm flexor muscles, trapezius muscles or erector spinae muscles, which are closely related to the sitting posture and vibration response, to obtain surface electromyography signals.

[0035] It should be noted that when the human body is subjected to continuous or periodic vibration, if the excitation frequency resonates with the local structure of the body (such as the spine and the head), it will significantly affect the steady state of the muscle system and the nervous system, causing fatigue, discomfort and even long-term health risks. Surface electromyography signals can sensitively represent changes in muscle activation caused by a vibrating environment, providing a direct and objective physiological basis for comfort judgment, thereby providing high-confidence feedback for adaptive adjustment of ride performance parameters such as suspension damping, energy recovery strength and driving mode.

[0036] S2, generating a vehicle comfort score based on the electromyography signals of all passengers.

[0037] In some embodiments, the surface electromyography signals of passengers in the vehicle are collected by the wearable device 200, and the collected electromyography signals are transmitted in real time to the vehicle infotainment system 100 through the wireless communication module built-in the wearable device 200; the vehicle infotainment system 100 analyzes and processes the received electromyography signals to generate a vehicle comfort score reflecting the overall feeling of the passengers. This scheme effectively solves the problem that existing vehicles lack real-time, direct and human physiological response-based detection means for ride smoothness or ride comfort during operation, realizes a technical leap from vehicle state perception to passenger state perception, and provides an objective and dynamic feedback basis for comfort control.

[0038] In another embodiment, when the wearable device 200 has sufficient computing power, the calculation process of the comfort score can be deployed locally on the wearable device 200, so that the complete calculation process from surface electromyogram signal acquisition, preprocessing to vehicle comfort score generation is completed inside the device; then, the wearable device 200 only needs to send the final comfort score result to the vehicle system 100 through the wireless communication module, without transmitting the original electromyogram signal data. This way not only reduces the computing burden of the vehicle system 100, but also reduces the data volume and power consumption of wireless transmission, while improving the system response speed and privacy protection level.

[0039] S3, adjusting the driving performance parameters of the vehicle according to the vehicle comfort score.

[0040] Specifically, the vehicle system 100 dynamically adjusts the driving performance parameters of the vehicle, including suspension damping, energy recovery intensity, and gear of the driving mode, according to the vehicle comfort score obtained in the previous step, so that the overall running state of the vehicle evolves towards higher comfort. This application takes the surface electromyogram signal of the occupant as the direct basis for comfort evaluation, and drives the real-time regulation of the driving performance parameters of the vehicle, realizing closed-loop feedback control from the human body feeling the behavior of the vehicle. Compared with the traditional method which only relies on passive adjustment of vehicle kinematic parameters or preset scene rules, this application can truly capture the muscle tension state of the occupant under different road conditions and driving conditions, so as to generate a comfort score in an objective, individualized and group integrated manner, and accordingly make collaborative, hierarchical and fine-tuned adaptive adjustments to multiple driving performance parameters, significantly improving the accuracy, initiative and adaptation ability of comfort control.

[0041] S4, after the adjustment of the driving performance parameters of the vehicle is completed, the system repeats steps S1 to S4 to form a closed-loop control process, so as to continuously collect the electromyogram signal of the occupant, update the comfort score and dynamically adjust the driving performance parameters, in order to maintain the comfort of the occupants in the vehicle in real time during driving.

[0042] It should be noted that if the driver manually intervenes and adjusts any driving performance parameter (such as switching the driving mode or modifying the suspension hardness) during operation, the system will pause automatic adjustment and follow the driver's setting; in addition, at the start of the next trip, the driving performance parameters of the vehicle will automatically restore to the standard default settings or the user's pre-saved custom configuration. Wherein, "trip" refers to a continuous driving process of the vehicle from the starting point to the destination, and the division basis is: when the vehicle is in the off state or idling state for 10 minutes or more, it is considered that the current trip is over, and the subsequent start is the beginning of a new trip. This mechanism not only guarantees the continuity of automatic comfort regulation, but also respects the initiative control right and individual preference of the driver.

[0043] In the above embodiments, the application realizes a complete technical closed loop of direct perception, quantitative evaluation and active optimization of ride comfort by combining real-time acquisition of occupant surface electromyography signals with dynamic adjustment of vehicle driving performance parameters. Compared with the prior art which relies on indirect indicators such as vehicle body vibration and steering angle change to infer smoothness, the application takes the electromyography signals of all occupants as the objective input source of comfort, directly reflecting the muscle tension state of the human body caused by bumps, jerks or power impact during vehicle operation, thereby truly and timely capturing the actual comfort perception of the occupants. On this basis, the system actively adjusts the driving performance parameters based on the generated vehicle comfort score to improve ride smoothness. This mechanism effectively overcomes the problems of adjustment lag, improper amplitude or wrong direction caused by the lack of human feedback in traditional methods, significantly improving the adaptive comfort protection capability of the vehicle under complex road conditions and diverse driving styles.

[0044] In addition, by fusing the electromyography signals of all occupants in the vehicle, the application avoids the limitations of focusing only on the driver or a single occupant, ensuring that the control decision is based on the comfort needs of the most sensitive occupant, and truly realizing a people-oriented intelligent comfort experience.

[0045] In some embodiments, the method of generating a vehicle comfort score includes: S21, generating different individual comfort scores based on the electromyography signals of each occupant.

[0046] Specifically, the wearable devices 200 worn by all occupants in the vehicle transmit the electromyography signals collected by them to the car system 100 through wireless transmission devices 210 such as Bluetooth or WIFI. After receiving the electromyography signals, the car system 100 analyzes and processes the electromyography signals of each occupant independently, generates corresponding individual comfort scores, and uses them for subsequent fusion calculation or control decision.

[0047] Among them, the individual comfort score is divided into four gears, and the comfort degree from low to high is represented by 0 to 3 points, and each score corresponds to different physiological and subjective feeling states. Specifically, 0 points represent that the occupant is continuously dizzy, accompanied by rapid heartbeat, difficulty breathing and other serious physiological stress reactions, indicating that the current vehicle operating state has a significant impact on the occupant's health; 1 point represents that the occupant appears dizzy and irritable, accompanied by nausea and other obvious symptoms, and the electromyography signal amplitude is significantly increased, reflecting that the muscle is in a tense or compensatory state; 2 points represent that the occupant has inattentive, slightly uncomfortable, and electromyography activity slightly increased but still tolerable; 3 points represent that the occupant has no abnormalities, the electromyography signal is stable, the attention is concentrated, and there is no discomfort. This scoring system combines surface electromyography signal features with typical human performance to provide a reliable basis for generating objective and quantifiable individual comfort evaluation.

[0048] S22, fuse the individual comfort scores of all passengers to obtain a vehicle comfort score.

[0049] Specifically, after generating the individual comfort scores of all passengers, the infotainment system 100 fuses all the individual comfort scores to obtain a vehicle comfort score that comprehensively reflects the overall feeling of the passengers in the vehicle. This score is used to represent the current overall comfort state of the vehicle and serves as a unified basis for subsequent adaptive adjustment of driving performance parameters.

[0050] In the above embodiments, by generating individual comfort scores for each passenger and fusing them, the application achieves comprehensive representation of the comfort state of all passengers in the vehicle, avoids control bias caused by relying solely on signals from a single passenger (such as the driver), and makes the vehicle control decision more global and fair.

[0051] In some embodiments, the method of generating individual comfort scores includes: S211, pre-process the electromyography signal to obtain an effective physiological signal.

[0052] Specifically, after obtaining the raw electromyography signal at the infotainment end, it needs to be pre-processed to extract the effective physiological signal. This pre-processing process mainly includes frequency domain filtering, aiming to suppress various noise interference and retain the effective components related to muscle activity. Specifically, the effective frequency band of surface electromyography signal is usually concentrated between 20Hz and 500Hz, so a band-pass filtering strategy is used for noise reduction.

[0053] It should be noted that the effective frequency band is limited to 20Hz to 500Hz, mainly because components below 20Hz mainly include baseline drift and motion artifacts caused by electrode displacement, slow limb movement or breathing, which do not carry true muscle electrical activity information and will mask the effective signal, affecting the stability of subsequent analysis; while high-frequency components above 500Hz are usually derived from environmental electromagnetic interference (such as vehicle electronic device radiation), power frequency harmonics or amplifier thermal noise, and the frequency spectrum energy of human muscle fiber action potential above 500Hz has sharply decayed, almost containing no physiological features for discrimination.

[0054] S212, based on the effective physiological signal, obtain a time sequence feature representing muscle state.

[0055] It should be noted that muscle state refers to the comprehensive embodiment of physiological responses such as muscle activation level, tension degree and fatigue trend of passengers in the vehicle caused by factors such as road bumps, vehicle acceleration changes or vibrations during vehicle driving; this state evolves dynamically over time and is directly reflected in the amplitude, frequency distribution and energy change characteristics of the surface electromyography signal.

[0056] Further, the time-series feature refers to a signal representation capable of quantifying the intensity, variation trend and response mode of muscle activity in the time dimension, which reflects the dynamic response of the passenger's muscle to external excitation (such as vibration, acceleration and deceleration) during vehicle driving; such a feature is obtained by time series analysis of the effective physiological signal, has continuity, stability and repeatability, and provides reliable and discriminative input basis for the subsequent comfort score model, thereby supporting continuous, objective and individualized assessment of the passenger's comfort state.

[0057] S213, generating a corresponding individual comfort score based on the time-series feature.

[0058] In the above embodiments, the present application filters out noise and interference by pre-processing the electromyography signal to obtain high-quality effective physiological signals, and further extracts time-series features capable of representing the muscle state of the passenger, so that the data input to the multi-classification model has clear physiological significance and time dynamic characteristics; this method effectively improves the consistency between the individual comfort score and the passenger's true subjective feeling, avoids misjudgment caused by directly using the original electromyography signal, and thereby provides a reliable basis for precise adjustment of vehicle driving performance parameters.

[0059] In some embodiments, the effective physiological signal is input to a time-series neural network to obtain a time-series feature; The time-series feature is input to a multi-classification model to obtain an individual comfort score.

[0060] In this embodiment, the multi-classification model is a classification head connected to the time-series neural network, and the time-series neural network and the classification head constitute an end-to-end model.

[0061] In some embodiments, the time-series neural network can adopt a deep learning architecture suitable for one-dimensional time series modeling, such as a long short-term memory network (LSTM), a gated recurrent unit (GRU), or a Transformer model based on a self-attention mechanism, to effectively capture the dynamic characteristics and long-term dependencies of the electromyography signal in the time dimension.

[0062] Further, the time-series neural network adopts a feedforward network composed of a fully connected layer, a batch normalization layer and an activation function for primary feature extraction, and then models the time series through a double-layer gated recurrent unit (GRU), and finally outputs the time-series feature of the electromyography signal in the time period as the input feature of the individual comfort score.

[0063] In some embodiments, the classification head is composed of one to two fully connected layers, and adopts a Softmax activation function in the output layer.

[0064] Specifically, the pre-processed effective physiological signal is first framed by a sliding window, that is, a continuous electromyographic signal of a fixed length d is taken as a frame, then the window is moved forward along the time axis with a sliding step s, and each subsequent frame is taken in turn, so that the whole signal is divided into a plurality of continuous data frames. The fixed length d refers to the signal length obtained according to the sliding step s, and can also be understood as a predefined time length.

[0065] Subsequently, each frame of electromyographic signal is inputted into a time sequence neural network. The network models the complete sequence in the frame, and takes the hidden state of the last time step as the output feature of the frame. The hidden state effectively fuses the dynamic changes and context-dependent information of the electromyographic signal in the time window, and can be used as a time sequence feature representing the current muscle tension and comfort state of the occupant.

[0066] Then, the time sequence feature corresponding to each frame is sent into a classification head respectively, and is mapped into a probability distribution of a plurality of preset comfort levels. The system selects the class with the highest probability as the individual comfort score corresponding to the time window. The score will be used for subsequent multi-occupant comfort fusion and real-time adjustment of vehicle driving parameters.

[0067] In addition, the time sequence neural network and the classification head jointly constitute a comfort evaluation model, and are jointly trained in an end-to-end manner to form a complete signal-to-semantic mapping system; the time sequence neural network is responsible for feature extraction in the time dimension of the input raw electromyographic signal, and captures the dynamic changes and context-dependent relationships of muscle activity in a continuous time period; the classification head maps the time sequence feature into a probability distribution of a plurality of preset discrete comfort levels, and outputs the class with the highest probability as the individual comfort score corresponding to the current time window. Through this joint modeling method, the system can directly learn discriminative representations highly related to subjective comfort perception from raw electromyographic signals, and realize efficient, accurate and low-delay comfort state prediction.

[0068] It is worth noting that the training method of the comfort evaluation model includes the following steps: Step 1, synchronously collecting physiological signals and subjective comfort labels.

[0069] Under various driving environments and road conditions (such as urban roads, highways, bumpy sections, etc.), the surface electromyographic signals of the driver and the occupants are collected by the wearable device 200; at the same time, the individual comfort scores of the subjects in the corresponding time period are obtained by using questionnaire survey, real-time voice feedback or on-site follow-up, etc. The comfort score is divided according to the preset level (for example, 0 to 3 points), which is used to represent the subjective feeling from severe discomfort to no difference.

[0070] Table 1 is an individual comfort score survey table

[0071] Table 1 Step 2, pre-process the original electromyography signal to extract the effective physiological signal.

[0072] The collected original electromyography signal is denoised, mainly including band-pass filtering to remove low-frequency baseline drift and high-frequency noise, and if necessary, signal normalization, detrending or power interference suppression, so as to obtain high-quality effective physiological signal.

[0073] Step 3, sliding window framing of the pre-processed signal.

[0074] The continuous electromyography signal is intercepted as a frame with a fixed length d, and the window is slid along the time axis with a preset sliding step s, so that the entire effective physiological signal is divided into a plurality of continuous and possibly overlapping data frames, each frame being an independent training sample.

[0075] Step 4, label completion for the time period with missing labels.

[0076] Since subjective scoring is usually a discrete event, some data frames may not directly correspond to an explicit comfort level label. To this end, the nearest neighbor interpolation method or other time alignment strategy is used to assign the subjective score of the adjacent period to the unlabeled frame, so as to assign the corresponding comfort level label to all data frames.

[0077] Step 5, training of the comfort level evaluation model based on the labeled data frames.

[0078] Each frame of electromyography signal and its corresponding comfort level label is combined into a training sample, which is input into the comfort level evaluation model. Through supervised learning, the model parameters are optimized so that it can automatically learn discriminative features related to subjective comfort from electromyography signals, and finally realize accurate prediction of individual comfort level.

[0079] Step 6, comfort level monitoring based on real-time electromyography signal in the test or actual operation stage.

[0080] The trained comfort level evaluation model is deployed to the vehicle system 100 or edge computing device; when the vehicle is running, the system receives real-time surface electromyography signals from the wearable device 200, which are pre-processed and framed by sliding window after the same pre-processing and sliding window framing, and then input into the model for inference. The current time window corresponding to the individual comfort level score is periodically output, and compared with the individual comfort level score fed back by the subject in the corresponding time period.

[0081] In the above embodiments, the application can automatically mine the nonlinear time dynamic patterns contained in the electromyographic signal, such as the instantaneous muscle reflex triggered by road vibration, the tension accumulation effect or fatigue evolution trend caused by continuous low-frequency vibration and other complex physiological responses, by inputting the effective physiological signal into the time sequence neural network and taking the high-dimensional representation of the output as the time sequence feature representing the muscle state. This method does not rely on artificial preset statistical features or domain prior knowledge, avoiding the neglect of dynamic details and information loss of traditional manual features. Further, the time sequence feature is input into the classification head associated with the time sequence neural network to directly generate a multi-grade individual comfort score, realizing an end-to-end mapping from the original physiological signal to the human factor evaluation. Compared with the machine learning method based on static features, this scheme significantly improves the system's perception sensitivity to subtle discomfort changes, the generalization ability between different road conditions and individual passengers, and the state tracking stability in continuous driving, thereby providing high-fidelity and low-delay comfort feedback basis for intelligent chassis to real-time and accurately adjust suspension damping, energy recovery strength and driving mode and other driving performance parameters.

[0082] In some embodiments, a feature value is calculated based on the effective physiological signal, and the feature value is taken as a time sequence feature. The time sequence feature is input into a multi-classification model to obtain an individual comfort score.

[0083] In this embodiment, the multi-classification model is implemented by a machine learning classifier.

[0084] The feature value includes at least one of a root mean square value, an integrated electromyographic value or a modified average absolute value.

[0085] The root mean square value reflects the muscle activation intensity by calculating the square root of the square mean value of the signal; the integrated electromyographic value represents the cumulative load of muscle activity by summing the absolute value of the signal; and the modified average absolute value improves the noise robustness by weighted average absolute value. The above feature calculation is simple and has clear physiological significance, which can effectively capture the muscle tension changes caused by vehicle vibration.

[0086] Specifically, the preprocessed effective physiological signal is first divided into frames by a sliding window, that is, a continuous electromyographic signal of a fixed length d is taken as a frame, then the window is moved forward along the time axis with a sliding step s, and each subsequent frame is sequentially taken, so that the entire signal is divided into a plurality of continuous data frames.

[0087] Subsequently, the feature value is independently calculated for each frame of signal to form a time sequence feature aligned with time; finally, the time sequence feature is input into the classifier, and the individual comfort score corresponding to the time window is output by the classifier.

[0088] Further, the classifier can be a machine learning model such as logistic regression (LR), support vector machine (SVM), or random forest.

[0089] In some embodiments, the feature values can further include at least one of waveform length, zero-crossing rate, average power frequency, sample entropy, and the like, in time domain, frequency domain, or non-linear dynamics, to more comprehensively characterize muscle activation intensity, fatigue state, and complexity of neuromuscular control, thereby improving the accuracy and environmental robustness of the comfort evaluation.

[0090] It is worth noting that the training method of the classifier is similar to the overall training process of the aforementioned comfort evaluation model (based on time series neural network), with the main difference being the fifth step: in the training of the classifier, the feature values extracted from each frame of electromyographic signal and their corresponding comfort labels are used as input samples to train a machine learning classifier such as logistic regression, support vector machine, or random forest, rather than an end-to-end neural network model.

[0091] In the above embodiments, the present application calculates classical time domain statistics such as root mean square, integral electromyographic value, or modified average absolute value from valid physiological signals as time series features representing muscle state, and inputs the time series features into a classifier to generate individual comfort scores, which not only significantly reduces the algorithmic complexity and memory usage, facilitating real-time processing on resource-constrained vehicle embedded platforms such as automotive-grade MCUs or low-power domain controllers, but also has clear physical meaning and interpretability, with stable mapping relationships between the numerical changes of the feature values and muscle activation intensity, sustained load level, and fatigue trend, facilitating calibration and alignment with controlled experiments and subjective comfort ratings (such as 0-3 points). In addition, compared to end-to-end modeling methods based on deep neural networks, this scheme significantly reduces the demand for training data size and computing resources, and the model training process is simple and efficient. Even under small sample conditions, a usable classifier can be quickly constructed through empirical thresholds or lightweight supervised learning, thereby significantly simplifying the system deployment, debugging, and functional verification process, while ensuring the reliability and cross-occupant generalization ability of the comfort scores, effectively reducing development costs, functional safety certification difficulty, and mass production landing threshold.

[0092] In some embodiments, the method of fusion processing is to calculate the mean or minimum value of the individual comfort scores of all occupants, and use the calculation result as the vehicle comfort score.

[0093] Specifically, the system first obtains individual comfort scores of all passengers (including the driver and passengers) wearing electromyography collection devices in the vehicle in a current time window in real time; then, the scores are aggregated according to a preset fusion strategy; if the mean value strategy is adopted, the arithmetic mean of all individual scores is calculated to reflect the overall average comfort level; if the minimum value strategy is adopted, the minimum value of all individual scores is selected to prioritize the riding experience of the least comfortable passenger. The aggregation result is the vehicle comfort score at the current time, which is used for subsequent coordinated adjustment of driving modes, suspension damping, or energy recovery intensity and other driving parameters.

[0094] In the above embodiments, the present application provides two optional solutions by calculating the mean value or minimum value of individual comfort scores of all passengers as the vehicle comfort score. The mean value can reflect the overall average comfort level of passengers in the vehicle, which is suitable for conventional scenarios that pursue comprehensive driving and riding quality. The minimum value takes the feeling of the least comfortable passenger as the basis for decision-making, effectively avoiding strong discomfort of individual passengers caused by road impact or vehicle dynamic response, and is particularly suitable for family travel or scenarios that focus on protecting vulnerable passengers (such as the elderly and children). In addition, the method is simple to calculate, has low resource overhead, and is easy to realize in real time in a vehicle controller, while taking into account system fairness and humanistic care, significantly improving the adaptability and user satisfaction of intelligent chassis adjustment.

[0095] In some embodiments, the driving performance parameters include suspension damping, energy recovery intensity, and driving mode, each of which is configured with at least two gears.

[0096] Specifically, the present embodiment adjusts the driving mode, suspension damping, and energy recovery strategy (including brake energy recovery and coasting energy recovery) in multiple dimensions based on the vehicle comfort score to reduce the adverse effects of vehicle vibration caused by road excitation or power system response on passenger comfort.

[0097] In the above embodiments, the present application sets at least two switchable gears for suspension damping, energy recovery intensity, and driving mode, respectively, and dynamically selects the target gear based on the vehicle comfort score, achieving fine and hierarchical adjustment of driving performance. Compared with the traditional fixed mode or single parameter adjustment method, this method can smoothly transition between multiple preset working conditions according to real physiological feedback of passengers, avoiding secondary discomfort caused by parameter mutation, and taking into account individual needs in different road conditions and passenger states, significantly improving the response accuracy, riding comfort, and user satisfaction of the intelligent chassis system.

[0098] In some embodiments, the method of adjusting the driving performance parameters of the vehicle includes: S31, mapping the vehicle comfort score to a preset comfort level interval; S32, set the suspension damping, energy recovery strength and / or driving mode gear according to the comfort level interval.

[0099] Specifically, the vehicle comfort score is divided into different comfort level intervals, each of which corresponds to a different adjustment strategy. The application adopts a rule-based control strategy to adjust the suspension damping, energy recovery strength and driving mode.

[0100] In addition, the number of adjusted driving performance parameters can vary for different comfort level intervals. Specifically, the suspension damping, energy recovery strength and driving mode gear can be adjusted simultaneously, or only one or two of them can be adjusted to achieve a control strategy that matches the current comfort needs.

[0101] In the above embodiments, the vehicle comfort score is mapped to a pre-set comfort level interval, and the corresponding parameter adjustment strategy is used according to the interval, so that the adjustment of the driving performance parameters can dynamically adapt to the change gradient of the passenger comfort state, avoiding the adjustment lag or over-response caused by a single fixed rule. The hierarchical mapping mechanism provides a structured decision-making framework for the system, which supports strong intervention measures in low comfort and allows conservative or maintenance strategies in high comfort, thereby improving the flexibility and robustness of the control system while ensuring the riding experience.

[0102] In some embodiments, the comfort level interval includes a first interval, a second interval, a third interval and a fourth interval, and the first interval to the fourth interval correspond to four levels of vehicle comfort score from low to high In some embodiments, the suspension damping, energy recovery strength and driving mode gear each include a first gear, a second gear and a third gear, and the first gear, the second gear and the third gear are arranged in order of comfort from high to low.

[0103] In some embodiments, when the vehicle comfort score is greater than or equal to 0 and less than or equal to 0.5, it is determined to be in the first interval; when the vehicle comfort score is greater than 0.5 and less than or equal to 1.5, it is determined to be in the second interval; when the vehicle comfort score is greater than 1.5 and less than or equal to 2.5, it is determined to be in the third interval; when the vehicle comfort score is greater than 2.5 and less than or equal to 3, it is determined to be in the fourth interval.

[0104] In some embodiments, the suspension damping is divided into three levels in order of comfort from high to low, namely comfort, standard and sport; wherein the comfort level corresponds to the first gear of the suspension damping, the comfort level is also the pre-set high comfort gear, the standard level corresponds to the second gear of the suspension damping, and the sport level corresponds to the third gear of the suspension damping.

[0105] In some embodiments, the energy recovery strength is divided into three levels: strong, medium and weak. Since the lower the energy recovery strength, the smoother the vehicle deceleration process, and the higher the passenger comfort, the order of comfort from high to low is weak, medium and strong; accordingly, weak recovery corresponds to the first gear of the energy recovery strength, weak recovery is also the preset high comfort gear, medium recovery corresponds to the second gear of the energy recovery strength, and strong recovery corresponds to the third gear of the energy recovery strength.

[0106] In some embodiments, the driving mode is divided into three levels according to the power response strength from high to low: sport mode, standard mode and economy mode. The weaker the power response, the softer the acceleration process, and the higher the comfort, so the order of comfort from high to low is economy, standard and sport; accordingly, the economy mode corresponds to the first gear of the driving mode, the economy mode is also the preset high comfort gear, the standard mode corresponds to the second gear of the driving mode, and the sport mode corresponds to the third gear of the driving mode.

[0107] It should be understood that for some vehicle models, the number of gears of suspension damping, energy recovery strength and driving mode can be expanded to more than three gears (for example, four or five gears) to achieve more fine adjustment of vehicle comfort, thereby better matching the individual needs of different passengers and diverse road conditions.

[0108] When the vehicle comfort score falls into the first interval, the passenger comfort is extremely poor at this time, so the gears of suspension damping, energy recovery strength and driving mode are all set to the first gear to quickly improve the vehicle comfort; that is, the suspension damping, energy recovery strength and driving mode are directly adjusted to the highest comfort gear; that is, the suspension damping is adjusted to the comfort gear, the energy recovery strength is adjusted to weak, and the driving mode is adjusted to the economy mode.

[0109] When the vehicle comfort score falls into the second interval, the gear change amount of suspension damping, energy recovery strength and driving mode is determined according to the first adjustment strategy, and the gears of the corresponding parameters are adjusted based on the determined gear change amount.

[0110] When the vehicle comfort score falls into the third interval, the gear change amount of suspension damping, energy recovery strength and driving mode is determined according to the second adjustment strategy, and the gears of the corresponding parameters are adjusted based on the determined gear change amount.

[0111] When the vehicle comfort score falls into the fourth interval, the comfort is basically as expected at this time, so the gears of suspension damping, energy recovery strength and driving mode remain unchanged.

[0112] In the above embodiment, the vehicle comfort score is divided into four preset interval, and different adjustment strategies are adopted for different intervals, so as to realize the hierarchical control of the driving performance parameters; in the second interval and the third interval, the system no longer adopts the traditional whole switching to the preset mode, but takes the gear change amount of each parameter as the decision variable, respectively calculates the suspension damping, the energy recovery strength and the gear number of the driving mode which should move to the high comfort direction, and adjusts the gear according to the demand, step by step and quantitatively, so as to avoid the excessive or insufficient adjustment caused by forced synchronous switching of all parameters; at the same time, by setting the gear with the highest comfort in the first interval and prohibiting the parameter disturbance in the fourth interval, the basic riding experience in the extreme uncomfortable scene and the system stability in the high comfort state are effectively guaranteed, and the adaptability and user perception consistency of the vehicle comfort control are improved.

[0113] In some embodiments, the first adjustment strategy includes a first target and a first condition.

[0114] In some embodiments, the first target is that the weighted comfort improvement benefit obtained by multiplying the gear change amount of the suspension damping by two, adding the gear change amount of the energy recovery strength by the gear change amount of the driving mode multiplied by 0.5, and subtracting the performance loss penalty calculated by multiplying the first coefficient by the sum of the gear change amount of the suspension damping, the gear change amount of the energy recovery strength and the gear change amount of the driving mode, the final result reaches the maximum value.

[0115] The first target can also be expressed by a formula, which is:

[0116] Further, the first condition is that only the suspension damping, the energy recovery strength and the driving mode are allowed to be adjusted to the high comfort gear direction, i.e. only adjustment from the third gear to the first gear direction is allowed, and at most only two of the suspension damping, the energy recovery strength and the driving mode are adjusted.

[0117] The first condition can also be expressed by a formula, which is:

[0118] represents the gear change amount of the suspension damping, represents the gear change amount of the energy recovery strength, represents the gear change amount of the driving mode; when adjusting from the third gear to the first gear direction, the gear change amount is positive; is the first coefficient; .

[0119] Further, ​is 0.1.

[0120] Further, when the first target and the first condition are simultaneously established, then the calculated gear change amount of the suspension damping the gear change amount of the energy recovery intensity the gear change amount of the driving mode The gears of the suspension damping, the energy recovery intensity, and the driving mode are adjusted respectively.

[0121] Further, when the first target and the first condition cannot be simultaneously established, it indicates that the optimal adjustment of the driving performance parameter has no solution, in order to avoid affecting the use, therefore the gears of the suspension damping, the energy recovery intensity, and the driving mode are directly set to the first gear, that is, the highest comfort gear.

[0122] In the above embodiment, the application adopts the first adjustment strategy of maximizing the weighted comfort benefit in the second interval, and limits the gear adjustment of at most only two driving performance parameters in the same control period, so that the system can moderately optimize the suspension damping, the energy recovery intensity, and the driving mode in the working condition where the passenger comfort is obviously insufficient but not extremely poor, and balance the comfort and the controllability; among them, the suspension damping is given the highest weight because it has the greatest impact on comfort, the energy recovery intensity is second, and the driving mode has the lowest weight, thereby prioritizing the chassis vibration filtering performance; in addition, when the optimization problem has no feasible solution, it automatically reverts to the preset comfort configuration, ensuring the robustness and safety of the control strategy.

[0123] In some embodiments, the second adjustment strategy includes a second target and a second condition.

[0124] Further, the second target is to maximize the weighted comfort degree improvement benefit obtained by multiplying the gear change amount of the suspension damping by two, adding the gear change amount of the energy recovery intensity by 0.5, and subtracting the performance loss penalty calculated by multiplying the first coefficient by the sum of the gear change amount of the suspension damping, the gear change amount of the energy recovery intensity, and the gear change amount of the driving mode.

[0125] Among them, the second target can also be expressed by the formula, which is specifically:

[0126] Further, the second condition is that only the suspension damping, the energy recovery intensity, and the driving mode are allowed to be adjusted to the high comfort gear direction, that is, only adjustment from the third gear to the first gear direction is allowed, and at most only one of the suspension damping, the energy recovery intensity, and the driving mode is adjusted.

[0127] The second condition can also be expressed by a formula, and is specifically:

[0128] The second condition can also be expressed by a formula, and is specifically: represents a gear change amount of the suspension damping, represents a gear change amount of the energy recovery intensity, represents a gear change amount of the driving mode; when the gear is adjusted from the third gear to the first gear, the gear change amount is positive; is a first coefficient; .

[0129] Further, is 0.1.

[0130] Further, when the second target and the second condition are simultaneously established, the gear of the suspension damping is adjusted according to the calculated gear change amount of the suspension damping , the gear of the energy recovery intensity is adjusted according to the calculated gear change amount of the energy recovery intensity , and the gear of the driving mode is adjusted according to the calculated gear change amount of the driving mode .

[0131] Further, when the second target and the second condition cannot be simultaneously established, it indicates that the optimization adjustment of the driving performance parameter has no solution, and in order to avoid affecting use, at this time the gears of the suspension damping, the energy recovery intensity and the driving mode are directly set to the first gear.

[0132] In the above embodiment, by using the same target function as the second interval but a more stringent parameter adjustment limit in the third interval, that is, only allowing gear adjustment of one driving performance parameter in the same control cycle, the system implements the minimum necessary intervention in the working condition where the passenger comfort is basically acceptable: preferentially selecting the single parameter (such as suspension damping) with the highest marginal benefit for current comfort improvement for fine tuning, while keeping the remaining parameters unchanged, thereby avoiding the sudden change of driving feeling or energy efficiency fluctuation caused by the synchronous change of multiple parameters; this strategy further optimizes the riding experience while maintaining the user's original driving style and vehicle dynamic characteristics; in addition, when there is no feasible solution to the optimization problem, it also falls back to the preset comfort configuration, thereby ensuring the control robustness in all working conditions.

[0133] In some embodiments, the comfort evaluation model no longer outputs a vehicle comfort score, but directly outputs target gears (such as the first gear, the second gear or the third gear) of the suspension damping, the energy recovery intensity and the driving mode. After the infotainment system 100 receives the set of target gears, the corresponding driving performance parameters are directly configured to realize comfort closed-loop control.

[0134] The training method of the model is similar to the overall process of the aforementioned comfort evaluation model based on time sequence neural network, but the training label is replaced by the calibrated optimal driving parameter combination instead of the subjective comfort score. Specifically, the actual impact of different driving performance parameter combinations on passenger comfort in various typical road conditions can be recorded by professionals, and the simultaneous collected electromyographic signals and feedback from the subjects are combined to determine the optimal gear combination corresponding to each physiological state. The optimal combination is used as a supervisory signal, paired with the corresponding electromyographic input frame, for end-to-end training of the neural network model, enabling it to directly predict the driving performance parameter configuration to be used from the original or feature-based electromyographic signals.

[0135] As shown in the accompanying Figures 5 to 7 The application also provides a driving performance parameter adjustment system applied to a vehicle, comprising: An electromyographic signal acquisition module, which is used to acquire electromyographic signals of all passengers in the vehicle; A signal preprocessing module 110 connected with the electromyographic signal acquisition module; the signal preprocessing module 110 is used to preprocess the electromyographic signals to obtain effective physiological signals; A comfort evaluation module 130 connected with the signal preprocessing module 110; the comfort evaluation module 130 is used to generate a vehicle comfort score based on the effective physiological signals of all passengers; A parameter adjustment execution module 140 connected with the comfort evaluation module 130; the parameter adjustment execution module 140 is used to adjust the driving performance parameters of the vehicle through the vehicle infotainment system 100 according to the vehicle comfort score to improve the comfort of the vehicle.

[0136] It is worth noting that the driver can choose whether to enable the function of the system in the default settings of the vehicle infotainment system 100.

[0137] In some embodiments, the electromyographic signal acquisition module includes an electromyographic bracelet worn on the wrist of the passenger. The electromyographic bracelet has at least two bioelectrodes built-in for contact with the skin to pick up electromyographic activity; in addition, a microcontroller and a low-power wireless communication module are integrated for data packaging and stable transmission. The raw electromyographic signals collected by the bracelet are transmitted to the vehicle infotainment system 100 of the vehicle itself via Bluetooth, Wi-Fi or other short-range wireless communication methods inside the device for subsequent analysis.

[0138] In some embodiments, the signal preprocessing module 110 includes a Butterworth filter. Specifically, to eliminate interference such as baseline drift and motion artifacts in the low frequency band, a Butterworth high-pass filter with a cutoff frequency of 20 Hz is used to perform high-pass filtering on the signal; then, to suppress high-frequency electromagnetic interference, power supply noise and other irrelevant high-frequency components, a Butterworth low-pass filter with a cutoff frequency of 500 Hz is used to perform low-pass filtering. Through the above two-stage filtering, a Butterworth band-pass filter of 20 Hz to 500 Hz is formed, which effectively retains the physiological information reflecting the muscle activation state, and significantly improves the signal-to-noise ratio of the signal to obtain effective physiological signals.

[0139] In some embodiments, the driving performance parameter adjustment system further includes a signal framing module 120 connected with the signal preprocessing module 110; the signal framing module 120 is configured to intercept continuous electromyographic signals with a fixed length d (e.g. 2 seconds to 10 seconds) as a frame, and slide the window along the time axis with a preset sliding step s (e.g. 1 second), to divide the entire effective physiological signal into a plurality of continuous and possibly overlapping data frames. Wherein, the function of the signal framing module 120 is mainly realized by a sliding window algorithm running on the car machine system 100.

[0140] In some embodiments, the comfort evaluation module 130 is connected with the signal framing module 120.

[0141] Further, the comfort evaluation module 130 includes a comfort evaluation model running on the microcontroller of the car machine system 100 or the wearable device 200.

[0142] Further, the comfort evaluation model includes a feature extraction sub-module and a classification head, and the feature extraction sub-module includes a pre-processing sub-module and a time series neural network sub-module. Wherein, the pre-processing sub-module is composed of a plurality of fully connected layers, batch normalization layers and ReLU activation functions in turn, for mapping the original electromyographic signal to a high-dimensional feature representation; the time series neural network sub-module adopts a double-layer stacked gated recurrent unit (GRU) structure, for context modeling in the time dimension of the feature sequence; and the classification head maps the final hidden state output by the time series modeling to a discrete comfort level.

[0143] Specifically, the comfort evaluation model first maps the single-channel raw electromyography signal through a fully connected layer with an input dimension of 1 and an output dimension of 16, and combines batch normalization and ReLU activation function to enhance the non-linear expression ability; then, the 16-dimensional features are further refined through the second-level fully connected layer (up to 64 dimensions) and the third-level fully connected layer (up to 128 dimensions), and the ReLU activation function is introduced after the first two levels of fully connected layers, and all three levels are equipped with batch normalization operation, so as to gradually convert the original time series signal into a high-dimensional and robust feature sequence. The 128-dimensional feature sequence is taken as the input of time series modeling and sent to the recurrent neural network stacked by two layers of GRU units, and the hidden state dimension of each layer of GRU is 128. The double-layer GRU processes the input sequence in time steps, dynamically updates its internal memory state at each time, and effectively captures the dynamic evolution law and long-range dependence of electromyography activity in time. After processing the signals in the entire time window, the hidden state output by the second layer of GRU at the last time step is taken as the final output of the time series neural network submodule, and the 128-dimensional vector condenses the comprehensive time series features of the electromyography signal in the current time period. Finally, the hidden state is input to the classification head, and the probability distribution of each comfort level is generated through the fully connected layer and the Softmax activation function, and then the classification prediction of individual comfort score is completed.

[0144] In the above embodiment, the present application realizes end-to-end control from passenger physiological state perception to adaptive adjustment of driving performance parameters by constructing a closed-loop control system composed of an electromyography signal acquisition module, a signal preprocessing module 110, a comfort evaluation module 130, and a parameter adjustment execution module 140: taking the electromyography signals of all passengers in the vehicle as the objective physiological basis for comfort evaluation, overcoming the limitations of traditional vehicle kinematics indicators such as vehicle speed change rate and steering angle fluctuation, which cannot truly reflect the muscle tension state of the human body; the effective physiological signals obtained after filtering out noise and interference by the signal preprocessing module 110 provide high signal-to-noise ratio input for the comfort evaluation module 130, significantly improving the accuracy of the comfort score; the parameter adjustment execution module 140 dynamically adjusts the suspension damping, energy recovery intensity, and driving mode based on the score, so that the vehicle can respond to the overall comfort needs of the passenger group in real time, actively optimize the riding experience without human intervention, and effectively improve the comfort of the vehicle.

[0145] Although the embodiments of the present application have been shown and described above, it can be understood that the above embodiments are exemplary and cannot be understood as limiting the present application, and those skilled in the art can make changes, modifications, replacements and variations to the above embodiments within the scope of the present application.

Claims

1. A method of adjusting a driving performance parameter, characterized by, The method is applied to a vehicle and comprises: collecting electromyography signals of all passengers in the vehicle; generating a vehicle comfort score based on the electromyography signals of all passengers; adjusting driving performance parameters of the vehicle according to the vehicle comfort score.

2. The method of adjusting a driving performance parameter according to claim 1, wherein The method for generating the vehicle comfort score comprises: generating individual comfort scores based on the electromyography signals of each passenger; fusing the individual comfort scores of all passengers to obtain the vehicle comfort score.

3. A method of adjusting a driving performance parameter according to claim 2, wherein The method for generating the individual comfort score comprises: preprocessing the electromyography signals to obtain effective physiological signals; obtaining time sequence features representing muscle states based on the effective physiological signals; generating a corresponding individual comfort score based on the time sequence features.

4. The method according to claim 3, wherein: the effective physiological signals are input into a time sequence neural network to obtain the time sequence features; or, feature values are calculated based on the effective physiological signals, and the feature values are taken as the time sequence features; the time sequence features are input into a multi-classification model to obtain the individual comfort score.

5. The method of adjusting a driving performance parameter according to claim 2, wherein The method for fusing comprises calculating a mean or minimum value of the individual comfort scores of all passengers, and taking the calculation result as the vehicle comfort score.

6. The method according to any one of claims 1 to 5, wherein The driving performance parameters include suspension damping, energy recovery intensity and driving mode; and the method for adjusting the driving performance parameters of the vehicle comprises: mapping the vehicle comfort score to a preset comfort level interval; setting gears of the suspension damping, the energy recovery intensity and / or the driving mode according to the comfort level interval.

7. A travel performance parameter adjustment method according to claim 6, characterized by, The comfort level interval comprises a first interval, a second interval, a third interval and a fourth interval; when the vehicle comfort score falls into the first interval, gears of the suspension damping, the energy recovery intensity and the driving mode are all set to their respective preset high-comfort gears; when the vehicle comfort score falls into the second interval, gear change amounts of the suspension damping, the energy recovery intensity and the driving mode are determined according to a first adjustment strategy, and gears of the corresponding parameters are adjusted based on the determined gear change amounts; when the vehicle comfort score falls into the third interval, gear change amounts of the suspension damping, the energy recovery intensity and the driving mode are determined according to a second adjustment strategy, and gears of the corresponding parameters are adjusted based on the determined gear change amounts; when the vehicle comfort score falls into the fourth interval, gears of the suspension damping, the energy recovery intensity and the driving mode remain unchanged.

8. A travel performance parameter adjustment method according to claim 7, characterized by, The first adjustment strategy comprises a first target and a first condition; the first target is that, a weighted comfort improvement benefit obtained by multiplying a gear change amount of the suspension damping by two, adding a gear change amount of the energy recovery intensity to the gear change amount of the driving mode multiplied by 0.5, and subtracting a performance loss penalty calculated by multiplying a first coefficient by a sum of the gear change amount of the suspension damping, the gear change amount of the energy recovery intensity and the gear change amount of the driving mode, reaches a maximum value. The first condition is that only the suspension damping, the energy recovery strength and the driving mode are allowed to be adjusted to the high comfort gear direction, and at most only two of the suspension damping, the energy recovery strength and the driving mode are adjusted; When the first target and the first condition are simultaneously established, the suspension damping, the energy recovery strength and the driving mode are adjusted according to the calculated gear change amount of the suspension damping, the gear change amount of the energy recovery strength and the gear change amount of the driving mode; When the first target and the first condition cannot be simultaneously established, the gears of the suspension damping, the energy recovery strength and the driving mode are all set to the high comfort gear.

9. A travel performance parameter adjustment method according to claim 7, characterized by, The second adjustment strategy includes a second target and a second condition; The second target is that the weighted comfort degree improvement benefit obtained by multiplying the gear change amount of the suspension damping by two, adding the gear change amount of the energy recovery strength and the gear change amount of the driving mode by 0.5, and subtracting the performance loss penalty calculated by multiplying the first coefficient by the sum of the gear change amount of the suspension damping, the gear change amount of the energy recovery strength and the gear change amount of the driving mode, reaches the maximum value; The second condition is that only the suspension damping, the energy recovery strength and the driving mode are allowed to be adjusted to the high comfort gear direction, and at most only one of the suspension damping, the energy recovery strength and the driving mode is adjusted; When the second target and the second condition are simultaneously established, the suspension damping, the energy recovery strength and the driving mode are adjusted according to the calculated gear change amount of the suspension damping, the gear change amount of the energy recovery strength and the gear change amount of the driving mode; When the second target and the second condition cannot be simultaneously established, the gears of the suspension damping, the energy recovery strength and the driving mode are all set to the high comfort gear.

10. A driving performance parameter adjustment system characterized by comprising: Applied to a vehicle, comprising: An electromyographic signal acquisition module, which is used to acquire electromyographic signals of all passengers in the vehicle; A comfort degree evaluation module (130), which is used to generate a vehicle comfort degree score based on the electromyographic signals of all passengers; A parameter adjustment execution module (140) connected with the comfort degree evaluation module (130); the parameter adjustment execution module (140) is used to adjust the driving performance parameters of the vehicle according to the vehicle comfort degree score.

Images