A method and system for predicting the degree of car sickness of an electric vehicle driver

Abstract

The application discloses a kind of electric vehicle driver and passenger car sickness degree prediction method and system, it is related to driver and passenger car sickness prediction field, the method includes obtaining historical driver and passenger physiological data and vehicle operation data, physiological data and vehicle operation data are preprocessed, the physiological data after preprocessing and the vehicle operation data after preprocessing are extracted, obtain the characteristic parameter related to car sickness, and the characteristic parameter related to car sickness is fused, and car sickness feature vector is obtained;Car sickness degree prediction model is constructed;Car sickness degree prediction model is trained, and the car sickness degree of the electric vehicle driver to be measured is predicted using the trained car sickness degree prediction model, to obtain car sickness degree prediction result, the application avoids the influence of subjective feeling feedback of driver and passenger in car sickness degree prediction process, and realizes the comprehensive and accurate prediction of driver and passenger car sickness degree.

Description

Technical Field

[0001] This application relates to the field of motion sickness prediction for drivers and passengers, and in particular to a method and system for predicting the degree of motion sickness in electric vehicle drivers and passengers. Background Technology

[0002] Motion sickness, a common form of motion sickness, primarily results from a conflict between the motion perceived by the inner ear's balance receptors and the visual information received during vehicle movement. This conflict disrupts the central nervous system, leading to a series of uncomfortable symptoms such as dizziness, nausea, vomiting, and paleness. Motion sickness not only significantly impacts the travel experience of drivers and passengers, reducing their satisfaction with electric vehicles, but can also pose a threat to driving safety in severe cases. With the continuous development and popularization of electric vehicles, their share in the transportation sector is increasing. While enjoying the many conveniences of electric vehicles, such as zero emissions, low noise, and lower operating costs, the occurrence of motion sickness among drivers and passengers has gradually attracted widespread attention.

[0003] In the field of traditional gasoline-powered vehicles, there are already many studies and countermeasures for motion sickness. However, due to the significant differences between electric vehicles and traditional gasoline-powered vehicles in terms of driving characteristics, powertrain layout, and vehicle vibration characteristics, these differences may have different effects on the motion sickness experience of drivers and passengers. For example, the power output of electric vehicles is more direct and rapid, and the vibration frequency and amplitude generated by its motor are different from those of an engine. This may change the human body's motion perception and balance state inside the vehicle. In addition, the center of gravity distribution, suspension system design, and interior space layout of electric vehicles may also be affected by factors such as the installation position of the battery pack, further affecting the comfort of drivers and passengers and their tendency to experience motion sickness.

[0004] Currently, existing methods for predicting motion sickness severity often rely primarily on subjective feedback from drivers and passengers, such as collecting descriptions of motion sickness symptoms and self-reported feelings during the journey through questionnaires. This approach not only suffers from significant subjectivity, making it difficult to accurately quantify the severity of motion sickness, but also fails to delve into the specific causes of motion sickness and the interactions between various influencing factors. Meanwhile, some studies have attempted to predict motion sickness status using physiological monitoring devices, such as heart rate variability monitors and electroencephalograms (EEGs), to indirectly reflect the state of motion sickness. However, these devices are typically complex and costly, and data interpretation requires specialized medical knowledge, making widespread adoption and application in actual vehicle design, development, and usage scenarios difficult. Summary of the Invention

[0005] The purpose of this application is to provide a method and system for predicting the degree of motion sickness of electric vehicle drivers and passengers, which can achieve comprehensive and accurate prediction of the degree of motion sickness of drivers and passengers, and avoid the problem of uncertainty in the degree of motion sickness due to the influence of subjective factors of drivers and passengers.

[0006] To achieve the above objectives, this application provides the following solution:

[0007] Firstly, this application provides a method for predicting the degree of motion sickness among electric vehicle drivers and passengers, including:

[0008] Acquire historical physiological data of drivers and passengers and vehicle operation data;

[0009] The physiological data and the vehicle operation data are preprocessed to obtain preprocessed physiological data and preprocessed vehicle operation data.

[0010] Feature extraction was performed on the preprocessed physiological data and the preprocessed vehicle operation data to obtain motion sickness-related feature parameters; the motion sickness-related feature parameters include: the average amplitude of the electroencephalogram (EEG) signal sequence. Mean Power spectral density of EEG signal sequence in the frequency domain PSD Wavelet transform of EEG signal sequence W Baseline level of skin conductance signal sequence SCL Power spectral density of skin conductance signal sequence in the frequency domain PSD S Wavelet transform sequence of skin conductance signals W S Number of eye fixations NOF Average eye fixation time MFD blinking frequency BR Facial expression key point distance d Facial muscle activity intensity MAI Vehicle acceleration variance and vehicle angular velocity variance ;

[0011] The motion sickness-related feature parameters are fused to obtain a motion sickness feature vector;

[0012] The motion sickness feature vectors are divided into a training set, a validation set, and a test set.

[0013] Construct a motion sickness severity prediction model;

[0014] The motion sickness severity prediction model is trained based on the training set to obtain a trained motion sickness severity prediction model.

[0015] A trained motion sickness prediction model was used to predict the degree of motion sickness among the drivers and passengers of the electric vehicle under test, and the prediction results of motion sickness were obtained.

[0016] Secondly, a motion sickness prediction system for electric vehicle drivers and passengers, the system comprising:

[0017] Wearable and fixed data acquisition devices are used to acquire historical physiological data of drivers and passengers and vehicle operation data.

[0018] An online motion sickness assessment platform, connected to the wearable and fixed data acquisition devices, is used to preprocess the physiological data and vehicle operation data to obtain preprocessed physiological data and preprocessed vehicle operation data. It is also used to extract features from the preprocessed physiological data and preprocessed vehicle operation data to obtain motion sickness-related feature parameters. These motion sickness-related feature parameters include the average amplitude of the electroencephalogram (EEG) signal sequence. Mean Power spectral density of EEG signal sequence in the frequency domain PSD Wavelet transform of EEG signal sequence W Baseline level of skin conductance signal sequence SCL Power spectral density of skin conductance signal sequence in the frequency domain PSD S Wavelet transform sequence of skin conductance signals W S Number of eye fixations NOF Average eye fixation time MFD blinking frequency BR Facial expression key point distance d Facial muscle activity intensity MAI Vehicle acceleration variance and vehicle angular velocity variance It is also used to perform feature fusion on the motion sickness-related feature parameters to obtain a motion sickness feature vector; it is also used to divide the motion sickness feature vector into a training set, a validation set, and a test set; it is also used to construct a motion sickness severity prediction model; it is also used to train the motion sickness severity prediction model based on the training set to obtain a trained motion sickness severity prediction model; it is also used to use the trained motion sickness severity prediction model to predict the motion sickness severity of the driver and passengers of the electric vehicle under test to obtain a motion sickness severity prediction result.

[0019] According to the specific embodiments provided in this application, this application has the following technical effects:

[0020] This application provides a method and system for predicting the severity of motion sickness in electric vehicle drivers and passengers. It involves acquiring historical physiological data of drivers and passengers and vehicle operation data, preprocessing the physiological data and vehicle operation data, extracting features from the preprocessed data to obtain motion sickness-related feature parameters. In this application, the motion sickness-related feature parameters are selected directly from the physiological data and vehicle operation data, avoiding the influence of subjective feedback from drivers and passengers during the entire motion sickness severity prediction process. The application also involves feature fusion of the motion sickness-related feature parameters to obtain a motion sickness feature vector, which is then divided into training data. This invention constructs a motion sickness severity prediction model using a set of data, a validation set, and a test set. The model is then trained to obtain a well-trained motion sickness severity prediction model. This model is then used to predict the motion sickness severity of passengers in the tested electric vehicle, thus solving the problem of uncertainty in motion sickness severity caused by subjective factors. This achieves comprehensive and accurate prediction of motion sickness severity. Furthermore, this application can directly output the motion sickness severity of passengers without requiring personnel with professional medical knowledge to interpret the physiological data of passengers and vehicle operation data, making it suitable for widespread use. Attached Figure Description

[0021] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0022] Figure 1 A flowchart illustrating a method for predicting motion sickness levels in electric vehicle drivers and passengers, provided in an embodiment of this application;

[0023] Figure 2 This is a schematic diagram of the functional modules of an electric vehicle driver and passenger motion sickness prediction system provided in an embodiment of this application. Detailed Implementation

[0024] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0025] The purpose of this invention is to provide a method and system for predicting the degree of motion sickness of electric vehicle drivers and passengers, so as to achieve a comprehensive and accurate prediction of the degree of motion sickness of drivers and passengers.

[0026] To make the above-mentioned objectives, features and advantages of this application more apparent and understandable, the application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0027] like Figure 1 As shown in the figure, this embodiment provides a method for predicting the degree of motion sickness of electric vehicle drivers and passengers, including the following steps:

[0028] Step 1: Obtain historical physiological data of drivers and passengers and vehicle operation data; the physiological data specifically includes: electroencephalogram (EEG) data, skin conductance data, eye movement data, and facial expression data.

[0029] Step 2: Preprocess the physiological data and vehicle operation data, including filtering, denoising, and normalization, to eliminate noise interference and dimensional differences in the data, improve the quality and usability of the data, and obtain preprocessed physiological data and preprocessed vehicle operation data.

[0030] Step 3: Extract features from the preprocessed physiological data and the preprocessed vehicle operation data to obtain feature parameters related to motion sickness.

[0031] Among the characteristic parameters related to motion sickness are: the average amplitude of the electroencephalogram (EEG) signal sequence. Mean Power spectral density of EEG signal sequence in the frequency domain PSD Wavelet transform of EEG signal sequence W Baseline level of skin conductance signal sequence SCL Power spectral density of skin conductance signal sequence in the frequency domain PSD S Wavelet transform sequence of skin conductance signals W S Number of eye fixations NOF Average eye fixation time MFD blinking frequency BR Facial expression key point distance d Facial muscle activity intensity MAI Vehicle acceleration variance and vehicle angular velocity variance The following section will elaborate on each feature parameter, including the following steps:

[0032] Average amplitude of EEG signal sequence Mean Power spectral density of EEG signal sequence in the frequency domain PSD and the EEG signal sequence after wavelet transform W The extraction process is as follows:

[0033] Step 301: Obtain an EEG signal sequence based on the preprocessed EEG data, and calculate the average amplitude of the EEG signal sequence. Mean Power spectral density of EEG signal sequence in the frequency domain PSD and the EEG signal sequence after wavelet transform W .

[0034] The average amplitude of the EEG signal sequence Mean It can be calculated using the following formula:

[0035] ;

[0036] in, This represents the first [number] in the EEG signal sequence. The amplitude of each EEG signal sampling point N This represents the total number of EEG signal sampling points. This feature reflects the average amplitude level of the EEG signal over a period of time.

[0037] The power spectral density of the frequency domain signal of the EEG signal sequence PSD Calculated in the following way:

[0038] First, the EEG signal sequence is subjected to a Fast Fourier Transform to convert it from the time domain to the frequency domain. FFT The result is a complex sequence containing the frequency components of the signal and their corresponding amplitudes, as shown in the following formula:

[0039] ;

[0040] in, x(n) It is the time-domain signal of the EEG signal sequence. It is a frequency domain signal of an electroencephalogram (EEG) signal sequence. It is the total number of sampling points for the electroencephalogram (EEG) signal. j It is the imaginary unit. k For frequency domain index, corresponding to the first in the discrete spectrum. k Spectral lines, , n This is the index of the time-domain sampling point, representing the first EEG signal on the time axis. n One sampling point, .

[0041] Secondly, based on the frequency domain signal of the EEG signal sequence, the power spectral density of the frequency domain signal of the EEG signal sequence is calculated. PSD The formula is as follows:

[0042] ;

[0043] in, Indicates to Take the mold. The power spectral density of the frequency domain signal of the EEG signal sequence is represented by calculation. FFT The power spectral density is obtained by squaredling the magnitude of the result and dividing by the number of sampling points. The power spectral density of the frequency domain signal of the EEG signal sequence represents the power distribution of the signal at different frequencies. It can help identify the main frequency components in the EEG signal, such as... α Wave (8-13Hz), β Wave (13-30Hz), etc.

[0044] The wavelet-transformed EEG signal sequence W It can be calculated using the following formula:

[0045] ;

[0046] in, It is a sequence of electroencephalogram (EEG) signals. These are wavelet basis functions. It is a scale parameter. These are translation parameters. Indicates conjugation. The result of wavelet transform. Wavelet coefficients, representing signals at different scales and locations, can be used to analyze the time-frequency characteristics of EEG signal sequences, such as detecting transient events and frequency changes in EEG signals. This application utilizes wavelet transform to simultaneously analyze EEG signals in both the time and frequency domains. By convolving the EEG signal with wavelet functions at different scales and locations, local features of the signal are extracted.

[0047] In addition to calculating the average amplitude of the above-mentioned electroencephalogram signal sequence, this embodiment also includes... Mean Power spectral density of EEG signal sequence in the frequency domain PSD and the EEG signal sequence after wavelet transform W In addition, based on the EEG signal sequence, the variance and standard deviation of the EEG signal sequence are calculated.

[0048] The variance of the EEG signal sequence is calculated using the following formula:

[0049] ;

[0050] in, The variance represents the magnitude of the EEG signal sequence. This variance measures the dispersion of the EEG signal amplitude. A larger variance indicates that the signal amplitude fluctuates more.

[0051] The standard deviation of the EEG signal sequence is calculated using the following formula:

[0052] ;

[0053] in, The standard deviation of the EEG signal sequence is the square root of the variance. It has the same dimensions as the original signal and more intuitively represents the magnitude of fluctuation in the EEG signal sequence.

[0054] This embodiment calculates the variance and standard deviation of the EEG signal sequence to determine whether the EEG signal sequence has quality abnormalities, providing information for subsequent process decisions. When the variance or standard deviation is too small, it indicates that the signal may be a straight line due to sensor detachment or poor contact during EEG signal acquisition. When the variance or standard deviation is too large, it indicates the possible inclusion of large-amplitude electromyography (EMG), electrooculography (EOG), or motion artifacts. Both small and large variances or standard deviations are marked as "abnormal signal quality." When the variance or standard deviation deviation is small, the EEG signal sequence is considered to have mild quality abnormalities, and it is given a low-weight quality label, which is automatically reduced during model training or inference. When the variance or standard deviation deviation is too large, the EEG signal sequence is considered to have severe quality abnormalities, and the entire segment is discarded or a re-acquisition is triggered.

[0055] In this embodiment, the feature extraction of EEG signals involves time-domain, frequency-domain, and time-frequency-domain analysis of EEG signal sequences, which can more comprehensively and accurately reflect the EEG status of drivers and passengers, making subsequent predictions of motion sickness severity more accurate.

[0056] Baseline level of skin conductance signal sequence SCL Power spectral density of skin conductance signal sequence in the frequency domain PSD S and the skin conductance signal sequence after wavelet transform W S The extraction process is as follows:

[0057] Step 302: Obtain the skin conductance signal sequence based on the preprocessed skin conductance data, and calculate the baseline level of the skin conductance signal sequence. SCL Power spectral density of skin conductance signal sequence in the frequency domain PSD S and the skin conductance signal sequence after wavelet transform W S .

[0058] The baseline level of the skin conductance signal sequence SCL It can be calculated using the following formula:

[0059] ;

[0060] in, This represents the first [number] in the skin conductance signal sequence. The amplitude of each skin conductance signal sampling point It is the total number of skin conductance signal sampling points. It reflects the baseline level of skin electrical conductance and is typically used for long-term trend analysis.

[0061] The power spectral density of the frequency domain signal of the skin conductance signal sequence PSD S Calculated in the following way:

[0062] First, a Fast Fourier Transform is performed on the skin conductance signal sequence, as shown in the following formula:

[0063] ;

[0064] in, y(n) It is the time-domain signal of the skin conductance signal sequence. It is the frequency domain signal of the skin conductance signal sequence. It is the total number of skin conductance signal sampling points. j It is the imaginary unit. k For frequency domain index, corresponding to the first in the discrete spectrum. k Spectral lines, , m The time-domain sampling point index represents the first sampling point in the skin conductance signal sequence. m One sampling point, .

[0065] Secondly, based on the frequency domain signal of the skin conductance signal sequence, the power spectral density of the frequency domain signal of the skin conductance signal sequence is calculated. PSD S The formula is as follows:

[0066] ;

[0067] in, To Take the mold. The power spectral density of the frequency domain signal of the skin conductance signal sequence PSD S .

[0068] The wavelet-transformed skin conductance signal sequence W S The following formula is used for calculation:

[0069] ;

[0070] in, It is a skin conductance signal sequence. (t) These are wavelet basis functions. It is a scale parameter. These are translation parameters. Indicates conjugate.

[0071] In addition to calculating the baseline level of the skin conductance signal sequence mentioned above, this embodiment also includes... SCL Power spectral density of skin conductance signal sequence in the frequency domain PSD S and the skin conductance signal sequence after wavelet transform W S In addition, based on the skin conductance signal sequence, the skin conductance response of the skin conductance signal sequence is calculated. SCR and skin conductance response SCR The frequency and amplitude.

[0072] The skin conductivity response SCR The following formula is used for calculation:

[0073] ;

[0074] It reflects the instantaneous changes in skin electrical conductance and is used for event correlation analysis.

[0075] The skin conductance response is calculated using the following formula. SCR Frequency and amplitude:

[0076] ;

[0077] ;

[0078] in, T The length of the time window, Indicates skin conductance response SCR The peak count can reflect the number of stress responses in skin conductance, i.e., the number of independent sympathetic activation events. Indicates skin conductance response SCR The frequency, which reflects the rate at which skin conductance events occur. This indicates the maximum value of the skin's electrical conductivity response. This represents the minimum value of the skin's electrical conductivity response. This indicates the amplitude of the skin conductance response, which reflects the activation intensity of the skin conductance response.

[0079] When the amplitude of skin conductance response If the value is too high or too low, it indicates that the patch of the skin conductivity acquisition device has fallen off or is saturated with sweat. In this case, the entire skin conductivity signal sequence data is marked as invalid. When the frequency of the skin conductivity response... When the waveform is too high and dense, it is suspected to be a motion artifact. Manual verification or independent component analysis (ICA) should be performed for further cleaning.

[0080] This embodiment calculates the skin conductance response based on the skin conductance signal sequence. SCR and skin conductance response SCR The frequency and amplitude of these parameters ensure that the samples entering the motion sickness prediction model are noise-free.

[0081] In this embodiment, the feature extraction of skin conductance signals involves time-domain, frequency-domain, and time-frequency-domain analysis of skin conductance signal sequences, which can comprehensively reflect the physiological state changes of drivers and passengers during electric vehicle operation, providing important physiological basis for assessing the degree of motion sickness.

[0082] Number of eye fixations in an eye movement signal sequence NOF Average eye fixation time MFD and blinking frequency BR The extraction process is as follows:

[0083] Step 303: Obtain an eye movement signal sequence based on the preprocessed eye movement data, and calculate the number of eye fixations in the eye movement signal sequence. NOF Average eye fixation time MFD and blinking frequency BR .

[0084] The number of eye fixation points NOF It can be calculated using the following formula:

[0085] ;

[0086] in, It represents the total number of fixations detected within a given time period. This feature reflects the number of eye fixations during a specific time interval.

[0087] The average fixation time of the eyes MFD The calculation is performed using the following formula:

[0088] ;

[0089] in, Indicates the first The duration of each fixation point. Average fixation time reflects how long the eyes linger at each fixation point.

[0090] blinking frequency BR The calculation is performed using the following formula:

[0091] ;

[0092] in, It is in time The number of blinks detected internally can reflect the blinking frequency of the eyes. A higher blinking frequency is related to the driver's or passenger's lack of concentration or fatigue.

[0093] In addition to calculating the number of eye fixation points mentioned above, this embodiment also includes... NOF Average eye fixation time MFD and blinking frequency BR In addition, based on the eye movement signal sequence, features are extracted from the eye movement signal trajectory, and the length and curvature of the eye movement trajectory are calculated.

[0094] The length of the eye movement trajectory is calculated using the following formula:

[0095] ;

[0096] in, Indicates the length of the eye movement trajectory. Indicates the first [item] on the eye movement trajectory The coordinates of each sampling point It is the total number of eye movement signal sampling points. This feature reflects the total length of the eye movement trajectory and can be used to assess the activity level of eye movements.

[0097] The curvature of the eye movement trajectory is calculated using the following formula:

[0098] ;

[0099] in, This represents the curvature of the eye movement trajectory. It refers to the angle change of the eye movement trajectory along a certain path. This corresponds to the arc length. The curvature of the eye movement trajectory reflects the degree of curvature of the eye movement trajectory; a higher curvature indicates a more complex eye movement trajectory.

[0100] When the eye movement track length is too short, it indicates that the sensor collecting the eye movement data is blocked, the driver or passenger closes their eyes, or other reasons have caused the eye movement signal to be lost. In this case, the entire eye movement signal sequence will be discarded or downweighted. When the curvature of the eye movement track suddenly increases, it indicates suspected blink artifacts or head movement interference. In this case, manual review or artifact removal process will be triggered.

[0101] This embodiment calculates the eye-track length and curvature, using them as hard thresholds during the data cleaning stage to ensure that subsequent training or inference uses noise-free samples.

[0102] This embodiment, through the special extraction of eye movement signals, can accurately reflect the eye movement of drivers and passengers, making the subsequent prediction of motion sickness more accurate.

[0103] Distance of facial expression key points in facial expression signal sequence d and the intensity of facial muscle activity MAIThe extraction process is as follows:

[0104] Step 304: Obtain a facial expression signal sequence based on the preprocessed facial expression data, and calculate the distance between facial expression key points in the facial expression signal sequence. d and the intensity of facial muscle activity MAI .

[0105] The distance of facial expression key points d Calculated in the following way:

[0106] First, obtain the coordinates of key facial expression points:

[0107] ;

[0108] in, Indicates the first The coordinates of key facial features It represents the total number of facial expression key points, and this coordinate can be used to describe the geometry of facial expressions.

[0109] Secondly, based on the aforementioned coordinates of key facial expression points, the distance between these key points is calculated. d The following formula is used:

[0110] ;

[0111] in, Indicates the first The key points of facial expressions and the first The distance between key facial expression points can be used to quantify changes in facial expressions, such as the degree of mouth opening or the degree of eyebrow raising.

[0112] The intensity of facial muscle activity MAI The following formula is used for calculation:

[0113] ;

[0114] in, Indicates the first The activity intensity of facial muscles, It is the total number of muscles, a feature that reflects the overall level of activity of the facial muscles of the driver and passengers.

[0115] In addition to calculating the distances to the aforementioned key facial expression points, this embodiment also includes... d and the intensity of facial muscle activity MAI In addition, the rate of change of facial muscle activity is calculated based on the facial expression signal sequence.

[0116] The rate of change in facial muscle activity is calculated using the following formula:

[0117] ;

[0118] in, This indicates the rate of change in facial muscle activity. This represents the change in the intensity of muscle activity. This represents the corresponding change over time; the rate of change in facial muscle activity can reflect the dynamic speed of change in facial expressions.

[0119] During the feature fusion stage, segments with high facial muscle activity change rates are assigned higher weights. This embodiment calculates the facial muscle activity change rate to ensure that the segments are not diluted by the static mean when the driver or passenger experiences sudden discomfort.

[0120] In this embodiment, the feature extraction of facial expression signals involves the distance between key facial expression points. d and the intensity of facial muscle activity MAI It can accurately capture changes in the facial expressions of drivers and passengers, making subsequent predictions of motion sickness severity more accurate.

[0121] Vehicle acceleration variance and vehicle angular velocity variance The extraction process is as follows:

[0122] Step 305: Calculate the vehicle acceleration variance based on the preprocessed vehicle operation data. and vehicle angular velocity variance .

[0123] The vehicle acceleration variance Calculated in the following way:

[0124] First, a vehicle acceleration signal sequence is obtained based on the preprocessed vehicle operation data, and the mean vehicle acceleration is calculated based on the vehicle acceleration signal sequence using the following formula:

[0125] ;

[0126] in, This represents the average vehicle acceleration. This represents the first [unit] in the vehicle acceleration signal sequence. Acceleration values ​​at each vehicle acceleration signal sampling point It is the total number of vehicle acceleration signal sampling points, which reflects the average acceleration level of the vehicle over a period of time.

[0127] Secondly, based on the above mean vehicle acceleration, the variance of vehicle acceleration is calculated. The following formula is used:

[0128] ;

[0129] In this embodiment, the vehicle acceleration variance It can measure the degree of fluctuation in vehicle acceleration. A larger variance indicates that the vehicle acceleration changes more significantly, which may be related to the frequent acceleration and deceleration of the vehicle.

[0130] The vehicle angular velocity variance Calculated in the following way:

[0131] First, a vehicle angular velocity signal sequence is obtained based on the preprocessed vehicle operation data, and the mean vehicle angular velocity is calculated based on the vehicle angular velocity signal sequence. The mean vehicle angular velocity is calculated using the following formula:

[0132] ;

[0133] in, This represents the average angular velocity of the vehicle. This represents the first [unit] in the vehicle angular velocity signal sequence. angular velocity values ​​at each vehicle angular velocity signal sampling point It is the total number of vehicle angular velocity signal sampling points, which reflects the vehicle's average rotational speed over a period of time.

[0134] Secondly, based on the mean vehicle angular velocity mentioned above, the variance of the vehicle angular velocity is calculated. The following formula is used:

[0135] ;

[0136] Variance of vehicle angular velocity in this implementation It can measure the degree of fluctuation in the vehicle's angular velocity. A larger variance indicates that the vehicle's rotational motion changes more significantly, which may be related to the vehicle's frequent turning.

[0137] In addition to calculating the aforementioned vehicle acceleration variance, this embodiment also includes... and vehicle angular velocity variance In addition, based on the preprocessed vehicle operation data, the maximum value of the vehicle acceleration signal sequence is calculated. and minimum value The maximum value of the vehicle angular velocity signal sequence and minimum value Average vehicle speed Vehicle speed variance and the maximum value of the vehicle speed signal sequence and minimum value .

[0138] The maximum value of the vehicle acceleration signal sequence and minimum value It can be calculated using the following formula:

[0139] ;

[0140] ;

[0141] The maximum value of the vehicle acceleration signal sequence and minimum value It can reflect the extreme acceleration and deceleration of a vehicle during driving.

[0142] The maximum value of the vehicle angular velocity signal sequence and minimum value It can be calculated using the following formula:

[0143] ;

[0144] ;

[0145] The maximum value of the vehicle angular velocity signal sequence and minimum value It can reflect the extreme rotational speed of a vehicle during operation.

[0146] Based on the preprocessed vehicle operation data, a vehicle speed signal sequence is obtained, and based on the vehicle speed signal sequence, the average vehicle speed is determined. It can be calculated using the following formula:

[0147] ;

[0148] in, This represents the first [unit] in the vehicle speed signal sequence. The speed values ​​of each vehicle speed signal sampling point It is the total number of vehicle speed signal sampling points, which reflects the average speed level of the vehicle over a period of time.

[0149] The vehicle speed variance It can be calculated using the following formula:

[0150] ;

[0151] The vehicle speed variance It measures the degree of fluctuation in vehicle speed. A larger variance indicates that the vehicle speed changes more significantly, which may be related to frequent acceleration and deceleration of the vehicle.

[0152] The maximum value of the vehicle speed signal sequence and minimum value It can be calculated using the following formula:

[0153] ;

[0154] ;

[0155] The maximum value of the vehicle speed signal sequence and minimum value It can reflect the extreme speed conditions of a vehicle during operation.

[0156] This embodiment calculates the maximum value of the vehicle acceleration signal sequence. and minimum value The maximum value of the vehicle angular velocity signal sequence and minimum value and the maximum value of the vehicle speed signal sequence and minimum value It can mark the timestamps of strong stimuli such as "rapid acceleration and deceleration, sharp turns"; and by aligning with physiological signals (SCR amplitude, rate of change of facial muscle activity) with delay, it can quantify the lag between physical stimulation and physiological response, helping the model learn the correct causal window.

[0157] In the feature fusion stage, this embodiment calculates the average vehicle speed. and vehicle speed variance Higher weights are given to physiological characteristics during periods of high variance to avoid diluting the model's attention with invalid data from low-speed, stable road sections.

[0158] In this embodiment, the extraction of vehicle motion features involves vehicle acceleration, angular velocity, and speed, which can accurately reflect the state of the vehicle used by the driver and passengers, making the subsequent prediction of motion sickness more accurate.

[0159] Step 4: Perform feature fusion on the motion sickness-related feature parameters to obtain a motion sickness feature vector, specifically including the following steps:

[0160] Step 401: Average amplitude of the EEG signal sequence Mean Power spectral density of EEG signal sequence in the frequency domain PSD and the EEG signal sequence after wavelet transform W We then perform weighted and fitted calculations to obtain the EEG feature vector.

[0161] The average amplitude of the electroencephalogram signal sequence Mean Power spectral density of EEG signal sequence in the frequency domain PSD and the EEG signal sequence after wavelet transform W We then perform weighted fitting using the following formula:

[0162] ;

[0163] in, The average amplitude of the EEG signal sequence Mean Power spectral density of EEG signal sequence in the frequency domain PSD and the EEG signal sequence after wavelet transform W Weighted and fitted EEG features , , , These are linear weighting coefficients, derived from linear least squares estimation. A 1 ，K 1 ，B 1 The parameters are nonlinear mapping parameters, obtained by quadratic fitting of the training set using nonlinear least squares.

[0164] The formula for the EEG feature vector is as follows:

[0165] ;

[0166] in, For EEG feature vectors, n To collect historical data on the number of drivers and passengers, The brainwave characteristics of the first driver and passenger. The EEG characteristics of the second passenger. For the first n Electroencephalogram (EEG) characteristics of the driver and passengers.

[0167] Step 402: Baseline level of the skin conductance signal sequence SCL Power spectral density of skin conductance signal sequence in the frequency domain PSD S and the skin conductance signal sequence after wavelet transform W S We then perform weighted summation and fitting to obtain the skin conductance feature vector.

[0168] Baseline level of the skin conductance signal sequence SCL Power spectral density of skin conductance signal sequence in the frequency domain PSD S and the skin conductance signal sequence after wavelet transform W S We then perform weighted fitting using the following formula:

[0169] ;

[0170] in, The baseline level of the skin conductance signal sequence SCLPower spectral density of skin conductance signal sequence in the frequency domain PSD S and the skin conductance signal sequence after wavelet transform W S Weighted and fitted EEG features b 1 , b 2 , b 3 , b 4 These are linear weighting coefficients, derived from linear least squares estimation. A 2 ，K 2 ，B 2 The parameters are nonlinear mapping parameters, obtained by quadratic fitting of the training set using nonlinear least squares.

[0171] The formula for the skin conductance feature vector is as follows:

[0172] ;

[0173] in, For skin conductance feature vectors, m To collect historical data on skin conductivity of passengers, The skin conductivity characteristics of the first passenger. The skin conductivity characteristics of the second passenger. For the first m The skin conductivity characteristics of the driver and passengers.

[0174] Step 403: Number of eye fixation points NOF Average eye fixation time MFD and blinking frequency BR We then perform weighted and fitted calculations to obtain the eye-tracking feature vector.

[0175] Number of eye fixation points NOF Average eye fixation time MFD and blinking frequency ​ We then perform weighted fitting using the following formula:

[0176] ;

[0177] in, The number of eye fixation points ​ Average eye fixation time ​ and blinking frequency ​ Weighted and fitted EEG features c 1, c 2 , c 3 These are linear weighting coefficients, derived from linear least squares estimation. A 3 ，K 3 ，B 3 ，C 3 The parameters are nonlinear mapping parameters, obtained by quadratic fitting of the training set using nonlinear least squares.

[0178] The formula for the eye-tracking feature vector is as follows:

[0179] ;

[0180] in, For eye-tracking feature vectors, p To collect historical driver and passenger data, The eye movement characteristics of the first driver / passenger. The eye movement characteristics of the second driver / passenger. For the first p Eye movement characteristics of drivers and passengers.

[0181] Step 404: Measure the distance between the key facial expression points. d and the intensity of facial muscle activity ​ We then perform weighted and fitted calculations to obtain facial expression feature vectors.

[0182] Distance of the key facial expression points d and the intensity of facial muscle activity ​ We then perform weighted fitting using the following formula:

[0183] ;

[0184] in, To determine the distance of the key facial expression points d and the intensity of facial muscle activity ​ Weighted and fitted facial expression features e 1 , e 2 , e 3 The coefficients are linear weighted and derived from linear least squares estimation. A 4 , K 4 , B 4 The parameters are nonlinear mapping parameters, obtained by quadratic fitting of the training set using nonlinear least squares.

[0185] The formula for the facial expression feature vector is as follows:

[0186] ;

[0187] in, It is a facial expression feature vector. q To collect historical data on facial expression of drivers and passengers, This refers to the facial expression characteristics of the first driver / passenger. The facial expression features of the second passenger. For the first q Facial expression characteristics of the driver and passengers.

[0188] Step 405: Calculate the variance of the vehicle acceleration. and vehicle angular velocity variance Weighted and fitted data are used to obtain the vehicle motion sickness feature vector.

[0189] The formula for the vehicle motion sickness feature vector is as follows:

[0190] ;

[0191] in, This is the feature vector of vehicle motion sickness. r To collect historical data on the number of drivers and passengers in vehicles, The motion sickness characteristics of the vehicle in which the first passenger was riding. For the first r Vehicle motion sickness characteristics of passengers in the vehicle.

[0192] Step 406: Weight and fit the EEG feature vector, skin conductance feature vector, eye movement feature vector, facial expression feature vector, and motion sickness feature vector to obtain a comprehensive feature vector. F .

[0193] The comprehensive feature vector F The formula is as follows:

[0194] .

[0195] Step 407: Process the comprehensive feature vector F The elements in the data are normalized to eliminate the influence of large differences in the dimensions and numerical ranges of different features on the fusion result. Normalization methods can include... ​ Normalization maps feature values ​​to the interval [0, 1]. For features The normalization formula is:

[0196] ;

[0197] in, and Representing features respectively Minimum and maximum values ​​among all samples This represents the normalized eigenvalues.

[0198] Step 408: Determine the sensitivity and contribution of the driver's motion sickness status based on the electroencephalogram (EEG) feature vector, the skin conductance feature vector, the eye movement feature vector, the facial expression feature vector, and the vehicle motion sickness feature vector. The weights can be determined through correlation analysis or machine learning methods; for example, principal component analysis (PCA) can be used to determine the weights of each feature.

[0199] Step 409: Based on the sensitivity and contribution of the driver and passengers' motion sickness status, determine the feature weights of the EEG feature vector, the skin conductance feature vector, the eye movement feature vector, the facial expression feature vector, and the vehicle motion sickness feature vector.

[0200] The formulas for the feature weights of the EEG feature vector, skin conductance feature vector, eye movement feature vector, facial expression feature vector, and motion sickness feature vector are as follows:

[0201] ;

[0202] ;

[0203] ;

[0204] ;

[0205] ;

[0206] in, The feature weights are the EEG feature vectors. The feature weights are the feature weights of the skin conductance feature vector. The feature weights are the feature weights of the eye-tracking feature vector. The feature weights of the facial expression feature vector. The feature weights are the feature weights of the vehicle motion sickness feature vector. For the first n The feature weights assigned to the EEG characteristics of each driver and passenger. For the first m The feature weights assigned to the skin conductivity characteristics of each driver and passenger. For the first pThe feature weights assigned to the eye movement characteristics of each driver and passenger. For the first q The feature weights assigned to the facial expression features of each driver and passenger. For the first r The feature weights assigned to the motion sickness characteristics of each driver and passenger in the vehicle.

[0207] Step 410: Multiply the normalized comprehensive feature vector by the feature weights of the EEG feature vector, the skin conductance feature vector, the eye movement feature vector, the facial expression feature vector, and the motion sickness feature vector to obtain a weighted feature vector. F weighted The weighted feature vector is the motion sickness feature vector.

[0208] The weighted feature vector F weighted The formula is as follows:

[0209] ;

[0210] in, F weighted represents a weighted eigenvector, and ⊙ represents element-wise multiplication. This represents the normalized EEG feature vector. This represents the normalized skin conductance feature vector. This represents the normalized eye-tracking feature vector. This represents the normalized facial expression feature vector. This represents the normalized vehicle motion sickness feature vector.

[0211] Through the above feature fusion process, this embodiment obtains a feature vector that comprehensively considers physiological signals and vehicle operating status. This feature vector can more comprehensively reflect the motion sickness state of drivers and passengers during the driving process of electric vehicles, providing strong data support for subsequent assessment of the degree of motion sickness and analysis of its causes.

[0212] Step 5: Divide the motion sickness feature vectors into a training set, a validation set, and a test set;

[0213] Step 6: Construct a motion sickness severity prediction model, which includes:

[0214] An input layer is used to receive the motion sickness feature vector.

[0215] A convolutional layer, connected to the input layer, includes convolutional kernels. Each convolutional kernel is used to extract local features of the motion sickness feature vector. Through convolution operations, the model can capture the spatial correlation in the feature vector.

[0216] A pooling layer, connected to the convolutional layer, is used to reduce the dimensionality of the motion sickness feature vector, reduce the amount of computation, and retain important feature information. The pooling operations in this application include max pooling and average pooling.

[0217] The LSTM layer, connected to the pooling layer, is used to learn the changes in motion sickness features over time. It can process time series data and capture the time dependence in the data.

[0218] A fully connected layer, connected to the LSTM layer, is used to perform weighted sum processing on the features output by the LSTM layer and map the processed features to a classification space for motion sickness severity.

[0219] The output layer is connected to the fully connected layer. The output layer uses the softmax activation function to convert the features output by the fully connected layer into a probability distribution, representing the probability that the driver and passengers are at different levels of motion sickness.

[0220] The motion sickness severity prediction model in this embodiment is a hybrid neural network model based on deep learning, specifically an architecture combining a convolutional neural network (CNN) and a long short-term memory network (LSTM). This model can effectively process time-series data and capture the spatial and temporal features of the data, thereby achieving an accurate assessment of the severity of motion sickness.

[0221] This embodiment can also utilize other machine learning or data mining algorithms, such as support vector machines and decision trees, to construct a motion sickness severity prediction model.

[0222] Step 7: Train the motion sickness severity prediction model based on the training set to obtain a trained motion sickness severity prediction model;

[0223] The training process specifically includes:

[0224] Step 701: Initialize the motion sickness prediction model constructed in Step 6: Initialize the model parameters, including convolutional kernel weights, LSTM unit parameters, and fully connected layer weights.

[0225] Step 702: Input the training set data into the model, perform forward propagation through convolutional layers, pooling layers, LSTM layers and fully connected layers, and calculate the model output.

[0226] Step 703: Use the cross-entropy loss function to calculate the difference between the model output and the true label. The cross-entropy loss function can measure the similarity between probability distributions and is suitable for multi-class classification problems.

[0227] Step 704: Based on the value of the loss function, calculate the gradient of the model parameters using the backpropagation algorithm, and update the model parameters using the Adam optimization algorithm.

[0228] In addition to training the motion sickness severity prediction model constructed in step 6, this embodiment also validates and tests the trained motion sickness severity prediction model based on the validation and test sets defined in step 5. Specifically, the model's performance is evaluated on the validation set, and the model parameters with the best performance are selected. The model's generalization ability is tested on the test set to ensure that the model can accurately assess the severity of motion sickness even on unseen data.

[0229] In this embodiment, to further improve the accuracy and generalization ability of the motion sickness severity prediction model, the following methods can be used for model evaluation and optimization:

[0230] ① Cross-validation: The motion sickness prediction model is evaluated using cross-validation to ensure its stability and consistency across different datasets.

[0231] ② Hyperparameter optimization: Optimize the hyperparameters of the motion sickness prediction model, such as learning rate, batch size, and convolutional kernel size, through methods such as grid search or random search, to improve the model's performance.

[0232] ③ Data augmentation: Enhance the training data, such as by adding noise or shifting the data, to increase the diversity of the data and improve the robustness of the motion sickness prediction model.

[0233] ④ Model ensemble: The Boosting model ensemble method is used to fuse the prediction results of multiple models to improve the overall prediction performance of the motion sickness prediction model.

[0234] Step 8: Use the trained motion sickness prediction model to predict the degree of motion sickness of the driver and passengers of the electric vehicle under test, and obtain the motion sickness prediction results.

[0235] In this embodiment, the motion sickness severity prediction results can be divided into different levels such as no motion sickness, mild motion sickness, moderate motion sickness, and severe motion sickness.

[0236] In this embodiment, the motion sickness feature vector, after feature extraction and fusion, is input into a trained neural network model. The model calculates the probability distribution of different levels of motion sickness experienced by the driver and passengers through forward propagation. The current level of motion sickness is determined based on the category corresponding to the maximum value of the probability distribution. For example, if the model outputs a probability distribution of [0.1, 0.3, 0.5, 0.1], it indicates that the driver and passengers have the highest probability of experiencing moderate motion sickness, and therefore, their motion sickness level is determined to be moderate.

[0237] Through the above-described method for predicting the degree of motion sickness among electric vehicle drivers and passengers, this embodiment can accurately and objectively predict the degree of motion sickness among drivers and passengers, and provide strong support for subsequent analysis of the causes of motion sickness and preventive measures.

[0238] Based on the aforementioned method for predicting motion sickness levels in electric vehicle occupants, this application also provides a system for predicting motion sickness levels in electric vehicle occupants. (See [link to relevant documentation]). ​ The electric vehicle passenger motion sickness prediction system includes:

[0239] Wearable and fixed data acquisition devices are used to acquire historical physiological data of drivers and passengers and vehicle operation data.

[0240] The wearable data acquisition device includes an electroencephalogram (EEG) and a skin conductance collector. The EEG is used to collect brainwave data and can detect changes in the brain's electrical activity. By analyzing the characteristics of brainwaves, such as frequency and amplitude, it indirectly reflects the brain's neural activity during motion sickness. The skin conductance collector is used to collect skin conductance data and can measure changes in the skin's electrical conductance. This indicator is closely related to the activity of the human autonomic nervous system. During motion sickness, skin conductance changes accordingly with the generation of emotions such as tension and anxiety, thus providing a physiological basis for evaluating motion sickness.

[0241] The fixed data acquisition equipment includes an eye tracker, a facial expression collector, and a gyroscope. The eye tracker is used to collect eye movement data, accurately recording the eye movement trajectory, fixation time, blink frequency, and other eye activity characteristics of the driver and passengers. Because the body's sense of balance and visual system are disturbed during motion sickness, the eye movement pattern will change. This eye movement data helps to analyze the impact of motion sickness on visual perception and eye movement. The facial expression collector is used to collect facial expression data. The facial expression collector can capture subtle changes in facial expressions. By recognizing and analyzing facial muscle movements and expression features, it can determine whether the driver and passengers are experiencing motion sickness-related discomfort expressions, such as frowning or opening their mouths. The gyroscope is used to collect vehicle operation data. In this embodiment, the gyroscope is mainly used to measure the rotational motion parameters of the vehicle during driving, including angular velocity and angular acceleration. This data can be used to analyze the impact of the vehicle's dynamic posture changes on the driver and passengers' experience of motion sickness.

[0242] The wearable data acquisition device also includes functional near-infrared spectroscopy (fNIRS), which uses near-infrared light to non-invasively monitor changes in blood oxygenation in the cerebral cortex, reflects the level of local neural activity in real time, and provides objective physiological indicators of brain function status for motion sickness prediction.

[0243] The changes in blood oxygenation in the cerebral cortex include changes in oxygenated hemoglobin concentration (HbO) and changes in deoxygenated hemoglobin concentration (HbR).

[0244] A motion sickness data acquisition and control terminal is provided, which is connected to the wearable acquisition device and the fixed acquisition device. It is used to perform basic communication operations such as formatting, packaging, and adding timestamps on the data collected by the wearable acquisition device and the fixed acquisition device. The data acquisition and control terminal establishes communication connections with the wearable acquisition device and the fixed acquisition device through a local area network and a 485 bus.

[0245] An online motion sickness assessment platform is connected to a motion sickness data acquisition and control terminal. This platform preprocesses physiological data and vehicle operation data after basic communication operations to obtain preprocessed physiological data and vehicle operation data. It also extracts features from the preprocessed physiological data and vehicle operation data to obtain motion sickness-related feature parameters. These motion sickness-related feature parameters include the average amplitude of an electroencephalogram (EEG) signal sequence. ​ Power spectral density of EEG signal sequence in the frequency domain ​ Wavelet transform of EEG signal sequence W Baseline level of skin conductance signal sequence ​ Power spectral density of skin conductance signal sequence in the frequency domain ​ S Wavelet transform sequence of skin conductance signals W S Number of eye fixations ​ Average eye fixation time ​ blinking frequency ​ Facial expression key point distance d Facial muscle activity intensity ​ Vehicle acceleration variance and vehicle angular velocity variance It is also used to perform feature fusion on the motion sickness-related feature parameters to obtain a motion sickness feature vector; it is also used to divide the motion sickness feature vector into a training set, a validation set, and a test set; it is also used to construct a motion sickness severity prediction model; it is also used to train the motion sickness severity prediction model based on the training set to obtain a trained motion sickness severity prediction model; it is also used to use the trained motion sickness severity prediction model to predict the motion sickness severity of the driver and passengers of the electric vehicle under test to obtain a motion sickness severity prediction result.

[0246] The motion sickness online evaluation platform establishes a communication connection with the motion sickness data acquisition and control terminal via Ethernet, mobile network, etc. It is responsible for receiving motion sickness-related data from various terminals, processing and analyzing the data using algorithms, generating motion sickness prediction results, and feeding back the results to the corresponding terminal devices, providing vehicle manufacturers, researchers, and drivers with an intuitive assessment of motion sickness.

[0247] The motion sickness assessment terminal APP connects to the online motion sickness assessment platform via 5G cellular communication. It is used to receive and display the motion sickness severity prediction results. It is installed on the mobile terminal device of the driver and passengers, such as a smartphone or tablet, so that the driver and passengers can record their motion sickness experience in real time while riding in an electric vehicle. This includes, but is not limited to, information such as the type, severity, and time of occurrence of motion sickness symptoms. The subjective experience data is then uploaded to the online motion sickness assessment platform to provide drivers and passengers with personalized motion sickness prevention suggestions and measures.

[0248] This embodiment relates to an electric vehicle passenger motion sickness prediction system. In practical application, firstly, fixed data acquisition devices, such as eye trackers, facial expression collectors, and gyroscopes, are installed and deployed inside the electric vehicle, ensuring a stable communication connection between them and the motion sickness data acquisition and control terminal via a local area network or RS-485 bus. Simultaneously, passengers wear wearable data acquisition devices, such as electroencephalograms (EEGs) and electrodermal data collectors, which are also connected to the motion sickness data acquisition and control terminal. When the electric vehicle begins to move, the fixed and wearable data acquisition devices synchronously start data acquisition, respectively acquiring vehicle operating status data and passenger physiological characteristic data, and transmit the data to the motion sickness data acquisition and control terminal in real time.

[0249] After processing and manipulating the received data, the motion sickness data acquisition and control terminal sends it to the motion sickness online evaluation platform via Ethernet or mobile network. Meanwhile, passengers can open the motion sickness evaluation terminal app at any time during the ride to record their motion sickness symptoms, such as the degree of dizziness and whether they feel nauseous, and submit them to the online motion sickness evaluation platform.

[0250] After receiving the aforementioned data, the online motion sickness assessment platform processes it according to a pre-set algorithm. First, the data undergoes preprocessing to remove noise and interference. Then, it extracts motion sickness-related feature parameters, such as power spectral density and characteristic frequencies from EEG data, skin conductance rate of change from ductus skinis data, and eye movement trajectory stability parameters from eye movement data. Next, these feature parameters are fused to construct a motion sickness feature vector, which is then input into a pre-trained machine learning model to obtain an assessment of the driver's / passenger's current level of motion sickness, for example, classifying it as mild motion sickness. Simultaneously, by combining vehicle operation data and the relationship between changes in various physiological feature parameters, the platform analyzes that the motion sickness may be due to frequent bumps and jolting on a particular road segment, leading to a loss of balance and thus triggering motion sickness symptoms.

[0251] The platform feeds back the motion sickness severity assessment results and motion sickness cause analysis report to the motion sickness evaluation terminal app. Drivers and passengers can use the app to view their specific motion sickness situation and causes, facilitating the implementation of corresponding preventative measures, such as adjusting seating posture and avoiding reading on bumpy roads. Furthermore, vehicle manufacturers can access the online motion sickness evaluation platform to collect a large amount of motion sickness data from drivers and passengers, conducting in-depth analysis of motion sickness under different vehicle models and driving conditions. This allows for targeted optimization of electric vehicle designs, such as improving suspension systems and adjusting power output characteristics, to reduce the probability of motion sickness and enhance the comfort and market competitiveness of electric vehicles.

[0252] This application presents a method and system for predicting motion sickness levels in electric vehicle drivers and passengers. It comprehensively considers the unique attributes of electric vehicles, accurately and objectively assesses the degree of motion sickness in drivers and passengers, and deeply analyzes the causes of motion sickness and related influencing factors. It provides strong technical support for electric vehicle manufacturers to optimize vehicle design, improve driving comfort, and ensure driving safety, and has application value and market prospects.

[0253] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0254] This document uses specific examples to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the methods and core ideas of this application. Furthermore, those skilled in the art will recognize that, based on the ideas of this application, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A method for predicting the degree of motion sickness in electric vehicle drivers and passengers, characterized in that, The method for predicting motion sickness levels in electric vehicle drivers and passengers includes: Acquire historical physiological data of drivers and passengers and vehicle operation data; the physiological data specifically includes: electroencephalogram (EEG) data, skin conductance data, eye movement data, and facial expression data; The physiological data and the vehicle operation data are preprocessed to obtain preprocessed physiological data and preprocessed vehicle operation data. Feature extraction is performed on the preprocessed physiological data and the preprocessed vehicle operation data to obtain feature parameters related to motion sickness. Specifically, this includes the following steps: Based on the preprocessed EEG data, the average amplitude of the EEG signal sequence was calculated. Mean Power spectral density of EEG signal sequence in the frequency domain PSD and the EEG signal sequence after wavelet transform W ; Based on the preprocessed skin conductance data, the baseline level of the skin conductance signal sequence was calculated. SCL Power spectral density of skin conductance signal sequence in the frequency domain PSD S and the skin conductance signal sequence after wavelet transform W S ; The number of eye fixations is calculated based on the preprocessed eye-tracking data. NOF Average eye fixation time MFD and blinking frequency BR ; Based on the preprocessed facial expression data, the distance between facial expression key points is calculated. d and the intensity of facial muscle activity MAI ; Based on the preprocessed vehicle operation data, the vehicle acceleration variance is calculated. and vehicle angular velocity variance ; The characteristic parameters related to motion sickness include: the average amplitude of the electroencephalogram (EEG) signal sequence. Mean Power spectral density of EEG signal sequence in the frequency domain PSD EEG signal sequence after wavelet transform W Baseline level of skin conductance signal sequence SCL Power spectral density of skin conductance signal sequence in the frequency domain PSD S Wavelet transform sequence of skin conductance signals W S Number of eye fixations NOF Average eye fixation time MFD blinking frequency BR Facial expression key point distance d Facial muscle activity intensity MAI Vehicle acceleration variance and vehicle angular velocity variance ; The motion sickness-related feature parameters are fused to obtain a motion sickness feature vector, specifically including the following steps: The average amplitude of the electroencephalogram signal sequence Mean Power spectral density of EEG signal sequence in the frequency domain PSD and the EEG signal sequence after wavelet transform W We then perform weighted and fitted calculations to obtain the EEG feature vector. Baseline level of the skin conductance signal sequence SCL Power spectral density of skin conductance signal sequence in the frequency domain PSD S and the skin conductance signal sequence after wavelet transform W S We then perform weighted and fitted calculations to obtain the skin electrical conductance feature vector. Number of eye fixation points NOF Average eye fixation time MFD and blinking frequency BR We then perform weighted and fitted calculations to obtain the eye-tracking feature vector. Distance of the key facial expression points d and the intensity of facial muscle activity MAI We then perform weighted and fitted calculations to obtain facial expression feature vectors. Regarding the variance of the vehicle acceleration and vehicle angular velocity variance Weighted and fitted data are used to obtain the vehicle motion sickness feature vector. The EEG feature vector, skin conductance feature vector, eye movement feature vector, facial expression feature vector, and motion sickness feature vector are weighted and fitted to obtain a comprehensive feature vector. F ; For the comprehensive feature vector F Normalize the elements in the data; The sensitivity and contribution of the driver and passenger's motion sickness status are determined based on the electroencephalogram feature vector, the skin conductance feature vector, the eye movement feature vector, the facial expression feature vector, and the vehicle motion sickness feature vector. Based on the sensitivity and contribution of the driver and passengers to motion sickness, the feature weights of the electroencephalogram feature vector, the skin conductance feature vector, the eye movement feature vector, the facial expression feature vector, and the vehicle motion sickness feature vector are determined. The normalized comprehensive feature vector is multiplied by the feature weights of the EEG feature vector, the skin conductance feature vector, the eye movement feature vector, the facial expression feature vector, and the motion sickness feature vector to obtain a weighted feature vector. F weighted The weighted feature vector is the motion sickness feature vector; The motion sickness feature vectors are divided into a training set, a validation set, and a test set. Construct a motion sickness severity prediction model; The motion sickness severity prediction model is trained based on the training set to obtain a trained motion sickness severity prediction model. A trained motion sickness prediction model was used to predict the degree of motion sickness among the drivers and passengers of the electric vehicle under test, and the prediction results of motion sickness were obtained.

2. The method for predicting the degree of motion sickness in electric vehicle drivers and passengers according to claim 1, characterized in that, The physiological data and the vehicle operation data are preprocessed, specifically by filtering, denoising, and normalizing them.

3. The method for predicting the degree of motion sickness in electric vehicle drivers and passengers according to claim 1, characterized in that, The formula for the EEG feature vector is as follows: ； in, For EEG feature vectors, n To collect historical data on the number of drivers and passengers, The brainwave characteristics of the first driver and passenger. The EEG characteristics of the second passenger. For the first n EEG characteristics of the driver and passengers; The formula for the skin electrical conductivity feature vector is as follows: ； in, For skin conductance feature vectors, m To collect historical data on skin conductivity of passengers, The skin conductivity characteristics of the first passenger. The skin conductivity characteristics of the second passenger. For the first m Skin conductivity characteristics of the driver and passengers; The formula for the eye-tracking feature vector is as follows: ； in, For eye-tracking feature vectors, p To collect historical driver and passenger data, The eye movement characteristics of the first driver / passenger. The eye movement characteristics of the second driver / passenger. For the first p Eye movement characteristics of the driver and passengers; The formula for the facial expression feature vector is as follows: ； in, It is a facial expression feature vector. q To collect historical data on facial expression of drivers and passengers, This refers to the facial expression characteristics of the first driver / passenger. The facial expression features of the second passenger. For the first q Facial expression characteristics of the driver and passengers; The formula for the vehicle motion sickness feature vector is as follows: ； in, This is the feature vector of vehicle motion sickness. r To collect historical data on the number of drivers and passengers in vehicles, The motion sickness characteristics of the vehicle in which the first passenger was riding. For the first r Vehicle motion sickness characteristics of passengers in the vehicle; The comprehensive feature vector F The formula is as follows: ； The formulas for the feature weights of the EEG feature vector, skin conductance feature vector, eye movement feature vector, facial expression feature vector, and motion sickness feature vector are as follows: ； ； ； ； ； in, The feature weights are the EEG feature vectors. The feature weights are the feature weights of the skin conductance feature vector. The feature weights are the feature weights of the eye-tracking feature vector. The feature weights of the facial expression feature vector. The feature weights are the feature weights of the vehicle motion sickness feature vector. For the first n The feature weights assigned to the EEG characteristics of each driver and passenger. For the first m The feature weights assigned to the skin conductivity characteristics of each driver and passenger. For the first p The feature weights assigned to the eye movement characteristics of each driver and passenger. For the first q The feature weights assigned to the facial expression features of each driver and passenger. For the first r The feature weights assigned to the vehicle motion sickness characteristics of each driver and passenger. The weighted feature vector F weighted The formula is as follows: ； in, F weighted represents a weighted eigenvector, and ⊙ represents element-wise multiplication. This represents the normalized EEG feature vector. This represents the normalized skin conductance feature vector. This represents the normalized eye-tracking feature vector. This represents the normalized facial expression feature vector. This represents the normalized vehicle motion sickness feature vector.

4. The method for predicting the degree of motion sickness in electric vehicle drivers and passengers according to claim 1, characterized in that, The motion sickness severity prediction model includes: Input layer, which is used to receive the motion sickness feature vector; A convolutional layer, connected to the input layer, the convolutional layer including convolutional kernels, each convolutional kernel being used to extract local features of the motion sickness feature vector; A pooling layer, connected to the convolutional layer, is used to reduce the dimensionality of the motion sickness feature vector; An LSTM layer, connected to the pooling layer, is used to learn how motion sickness characteristics change over time. A fully connected layer, connected to the LSTM layer, is used to perform weighted sum processing on the features output by the LSTM layer and map the processed features to a classification space for motion sickness severity. An output layer, connected to the fully connected layer, uses a softmax activation function to convert the features output by the fully connected layer into a probability distribution.

5. A motion sickness prediction system for electric vehicle drivers and passengers, characterized in that, The electric vehicle passenger motion sickness prediction system includes: Wearable and fixed data acquisition devices are used to acquire historical physiological data of drivers and passengers and vehicle operation data; the physiological data specifically includes: electroencephalogram (EEG) data, skin conductance data, eye movement data, and facial expression data. An online assessment platform for motion sickness, which is connected to the wearable data acquisition device and the fixed data acquisition device, is used to preprocess the physiological data and the vehicle operation data to obtain preprocessed physiological data and preprocessed vehicle operation data. The online motion sickness evaluation platform is also used to extract features from the preprocessed physiological data and preprocessed vehicle operation data to obtain motion sickness-related feature parameters. Specifically, this includes the following steps: calculating the average amplitude of the EEG signal sequence based on the preprocessed EEG data. Mean Power spectral density of EEG signal sequence in the frequency domain PSD and the EEG signal sequence after wavelet transform W Based on the preprocessed skin conductance data, the baseline level of the skin conductance signal sequence was calculated. SCL Power spectral density of skin conductance signal sequence in the frequency domain PSD S and the skin conductance signal sequence after wavelet transform W S Based on the preprocessed eye-tracking data, the number of eye fixations is calculated. NOF Average eye fixation time MFD and blinking frequency BR Based on the preprocessed facial expression data, the distance between facial expression key points is calculated. d and the intensity of facial muscle activity MAI Based on the preprocessed vehicle operation data, calculate the vehicle acceleration variance. and vehicle angular velocity variance The characteristic parameters related to motion sickness include: the average amplitude of the electroencephalogram (EEG) signal sequence. Mean Power spectral density of EEG signal sequence in the frequency domain PSD EEG signal sequence after wavelet transform W Baseline level of skin conductance signal sequence SCL Power spectral density of skin conductance signal sequence in the frequency domain PSD S Wavelet transform sequence of skin conductance signals W S Number of eye fixations NOF Average eye fixation time MFD blinking frequency BR Facial expression key point distance d Facial muscle activity intensity MAI Vehicle acceleration variance and vehicle angular velocity variance ; The online motion sickness evaluation platform is also used to perform feature fusion on the motion sickness-related feature parameters to obtain a motion sickness feature vector, specifically including the following steps: averaging the amplitude of the electroencephalogram (EEG) signal sequence... Mean Power spectral density of EEG signal sequence in the frequency domain PSD and the EEG signal sequence after wavelet transform W Weighted and fitted data are used to obtain the EEG feature vector; the baseline level of the skin conductance signal sequence is then calculated. SCL Power spectral density of skin conductance signal sequence in the frequency domain PSD S and the skin conductance signal sequence after wavelet transform W S Weighted and fitted data are used to obtain the skin conductance feature vector; the number of eye fixation points is then calculated. NOF Average eye fixation time MFD and blinking frequency BR Weighted and fitted data are used to obtain eye-tracking feature vectors; the distances to the key facial expression points are then calculated. d and the intensity of facial muscle activity MAI Weighted and fitted data are used to obtain facial expression feature vectors; the variance of vehicle acceleration is then analyzed. and vehicle angular velocity variance Weighted and fitted data are used to obtain the vehicle motion sickness feature vector; the EEG feature vector, skin conductance feature vector, eye movement feature vector, facial expression feature vector, and vehicle motion sickness feature vector are then weighted and fitted to obtain a comprehensive feature vector. F ; for the comprehensive feature vector F The elements in the vector are normalized; the sensitivity and contribution of the driver's motion sickness state to the EEG feature vector, skin conductance feature vector, eye movement feature vector, facial expression feature vector, and vehicle motion sickness feature vector are determined; based on the sensitivity and contribution of the driver's motion sickness state to the EEG feature vector, skin conductance feature vector, eye movement feature vector, facial expression feature vector, and vehicle motion sickness feature vector are determined; the normalized comprehensive feature vector is multiplied by the feature weights of the EEG feature vector, skin conductance feature vector, eye movement feature vector, facial expression feature vector, and vehicle motion sickness feature vector to obtain a weighted feature vector. F weighted The weighted feature vector is the motion sickness feature vector; The online motion sickness evaluation platform is also used to divide the motion sickness feature vector into a training set, a validation set, and a test set. The online motion sickness assessment platform is also used to build a motion sickness severity prediction model; The motion sickness online evaluation platform is also used to train the motion sickness severity prediction model based on the training set to obtain a trained motion sickness severity prediction model. The online motion sickness assessment platform is also used to predict the degree of motion sickness of the drivers and passengers of the electric vehicle under test using a trained motion sickness prediction model, and obtain the motion sickness prediction result.

6. The motion sickness prediction system for electric vehicle drivers and passengers according to claim 5, characterized in that, The motion sickness prediction system for electric vehicle drivers and passengers also includes a motion sickness evaluation terminal APP, which is connected to the online motion sickness evaluation platform and is used to receive and display the motion sickness prediction results.

7. The motion sickness prediction system for electric vehicle drivers and passengers according to claim 5, characterized in that, The wearable data acquisition device includes: an electroencephalogram (EEG) device and a skin conductance acquisition device; The electroencephalometer is used to collect electroencephalogram (EEG) data; The skin conductance acquisition device is used to collect skin electrical conductivity data; The fixed data acquisition device includes: an eye tracker, a facial expression collector, and a gyroscope; The eye tracker is used to collect eye movement data; The facial expression collector is used to collect facial expression data; The gyroscope is used to collect vehicle operation data.

Images