A method and system for evaluating driver fatigue

Abstract

The application discloses a kind of intelligent driving fatigue evaluation method and system, it is related to automobile safety technical field, method includes: obtaining the physiological data of driver in driving scene;The physiological data includes electromyographic data, heart rate data and respiratory rate data;According to the physiological data, calculate driving fatigue evaluation index;The driving fatigue evaluation index includes physical load, vigilance and cognitive load;According to driving fatigue evaluation index, using weight method calculates fatigue driving comprehensive evaluation index.The present application comprehensively considers psychological fatigue and physiological fatigue, from physical load, vigilance and cognitive load three dimensions, comprehensively electromyographic signal root mean square value, electromyographic signal integral value, heart rate mean value, the root mean square value of difference between adjacent NN intervals in whole process, heart rate variability low frequency power and high frequency power ratio and heart rate respiratory cross consistency and so on Index, using weight method carries out driving fatigue evaluation, can effectively improve the accuracy, objectivity of driving fatigue evaluation.

Description

Technical Field

[0001] This invention relates to the field of automotive safety technology, and in particular to an intelligent driving fatigue assessment method and system. Background Technology

[0002] Road traffic accidents pose a significant threat to life safety, resulting in numerous casualties and substantial property losses. Driver fatigue is a major contributing factor to road traffic accidents. While intelligent driving is being implemented, due to limited attention paid to intelligent cockpits, a standardized evaluation system and assessment method for fatigue driving during driving has not been established. Fatigue during driving is mainly divided into two categories: (1) Psychological fatigue: mental exhaustion caused by prolonged monotonous, repetitive work or sustained high-intensity cognitive load. (2) Physiological fatigue: decreased functional ability, manifested as muscle soreness, fatigue, and weakness. Driving fatigue is typically a mixed fatigue resulting from both psychological and physiological factors.

[0003] Currently, existing driver fatigue detection or evaluation technologies rely on various methods. Some rely on external behavioral characteristics, such as blinking, head-down movements, and hand grip strength; others rely on vehicle driving conditions, such as speed and lateral deviation. However, these methods struggle to directly and accurately understand the driver's psychological and physiological characteristics, resulting in low accuracy in driver fatigue detection. Still others rely on physiological signals, such as electroencephalogram (EEG) signals and heart rate, but these are largely based on individual physiological signals and lack comprehensive consideration, leading to continued low accuracy in driver fatigue evaluation. Therefore, a complete and scientific evaluation index system and effective evaluation methods are urgently needed to evaluate driver fatigue in intelligent cockpits. Summary of the Invention

[0004] The purpose of this invention is to provide an intelligent driving fatigue evaluation method and system that evaluates driving fatigue from two dimensions: psychological fatigue and physiological fatigue, so as to achieve accurate and objective evaluation.

[0005] To achieve the above objectives, the present invention provides the following solution:

[0006] In a first aspect, the present invention provides a method for evaluating fatigue in intelligent driving, the method comprising:

[0007] Acquire physiological data of the driver in a driving scenario; the physiological data includes electromyography data, heart rate data, and respiratory rate data;

[0008] Driving fatigue evaluation indicators are calculated based on the physiological data. These indicators include physiological and psychological driving fatigue evaluation indicators. The physiological indicators include physical exertion indicators, which include the root mean square value of electromyography (EMG) signals and the integral value of EMG signals. The psychological indicators include alertness and cognitive load indicators. The alertness indicators include the mean heart rate, the root mean square value of the difference between adjacent NN intervals, and the ratio of low-frequency to high-frequency power of heart rate variability, where the NN interval represents the interval between two adjacent heartbeats. The cognitive load indicators include the ratio of low-frequency to high-frequency power of heart rate variability and heart rate-respiratory cross-consistency.

[0009] Based on the driver fatigue evaluation indicators, the weighted method is used to calculate the comprehensive fatigue driving evaluation index.

[0010] Optionally, the formula for calculating the integral value of electromyography (iEMG) is as follows:

[0011]

[0012] Where Fs is the sampling frequency, EMG i This represents the amplitude of the i-th electromyographic signal, and N is the number of electromyographic signal amplitude data.

[0013] Optionally, the ratio of low-frequency power to high-frequency power of heart rate variability (LF / HF) can be calculated using the following formula:

[0014] LF = Sum(FFT(RR)) | 0.04～0.15Hz ,

[0015] HF=Sum∈(FFT(RR))| 0.15～0.4Hz ,

[0016] Where LF is the low-frequency power of heart rate variability, HF is the high-frequency power of heart rate variability, Sum() represents the summation operation, FFT() represents the fast Fourier transform operation, and RR is the heartbeat interval.

[0017] Optionally, the heart rate-respiration cross-consistency coefficient (CPC) can be calculated using the following formula:

[0018]

[0019] Where HR is heart rate, Resp is respiratory rate, Cross() calculates the cross-correlation between signals, and Coherence() calculates the coherence between signals.

[0020] Optionally, based on the driver fatigue evaluation indicators, a weighted method is used to calculate the comprehensive fatigue driving evaluation index, specifically including:

[0021] The fatigue driving comprehensive evaluation index A is calculated using the following formula:

[0022] A=μ1A1+μ2A2+μ3A3+μ4A4+μ5A5+μ6A6,

[0023] Where A1 is the root mean square value of the electromyography (EMG) signal, A2 is the integral value of the EMG signal, A3 is the root mean square value of the difference between adjacent NN intervals throughout the entire process, A4 is the ratio of low-frequency power to high-frequency power of heart rate variability, A5 is the mean heart rate, and A6 is the heart rate-respiration cross-consistency coefficient; μ1, μ 12 ,…,μ6 are the weight coefficients of A1, A2,…,A6 respectively.

[0024] Optionally, the formula for calculating the weighting coefficients μ1-μ6 is as follows:

[0025]

[0026] Where, μ i Let A be the i-th driver fatigue evaluation index. i Weighting coefficients; W i The weight of the i-th driver fatigue evaluation index determined by the analytic hierarchy process; w i is the weight of the i-th driving fatigue evaluation index determined by the entropy weight method; N is the total number of driving fatigue evaluation indicators.

[0027] Optionally, the process of determining the weight coefficient of the i-th driving fatigue evaluation index using the analytic hierarchy process is as follows:

[0028] The relative importance of each driver fatigue evaluation indicator was scored using the priority relation quantitative scaling method, and a judgment matrix was obtained.

[0029] The judgment matrix is ​​normalized column-wise, and the normalized judgment matrix is ​​then summed row-wise to obtain the sum vector M. i , where i represents the i-th driver fatigue evaluation index;

[0030] The weight coefficient W of the i-th driving fatigue evaluation index is calculated according to the following formula. i :

[0031]

[0032] Optionally, before performing the column-wise normalization of the judgment matrix, the following steps are also included:

[0033] If the number of driving fatigue evaluation indicators is greater than 2, then the judgment matrix is ​​subjected to a consistency check. If the judgment matrix fails the consistency check, then the judgment matrix is ​​modified and the consistency check is performed again until the judgment matrix passes the consistency check.

[0034] If the number of driving fatigue evaluation indicators is less than or equal to 2, then proceed to the step of normalizing the judgment matrix column by column.

[0035] Optionally, the process of determining the weight coefficient of the i-th driving fatigue evaluation index using the entropy weight method is as follows:

[0036] The raw data of each driver fatigue evaluation indicator are standardized to obtain unit data for each indicator. The information entropy and information utility values ​​of each indicator are then calculated using the following formulas:

[0037]

[0038] d i =1-e i ,

[0039] Among them, e i Let d be the information entropy value of the i-th driving fatigue evaluation index. i Let m be the information utility value of the i-th driver fatigue evaluation index. * Y represents the number of raw data points for a single driver fatigue evaluation indicator. ti This represents the t-th unit of data for the i-th driving fatigue evaluation index.

[0040] The weight coefficient w of the i-th driving fatigue evaluation index determined by the entropy weight method is calculated using the following formula. i :

[0041]

[0042] Secondly, the present invention provides an intelligent driving fatigue assessment system, comprising:

[0043] The physiological data acquisition module is used to acquire the driver's physiological data, including electromyography data, heart rate data, and respiratory rate data.

[0044] A driver fatigue evaluation index calculation module is used to calculate driver fatigue evaluation indices based on the physiological data. These indices include physiological driver fatigue evaluation indices and psychological driver fatigue evaluation indices. The physiological driver fatigue evaluation indices include physical load indices, which include the root mean square value of electromyography (EMG) signals and the integral value of EMG signals. The psychological driver fatigue evaluation indices include alertness indices and cognitive load indices. The alertness indices include the mean heart rate, the root mean square value of the difference between adjacent NN intervals throughout the journey, and the ratio of low-frequency power to high-frequency power of heart rate variability, where the NN interval represents the interval between two adjacent heartbeats. The cognitive load indices include the ratio of low-frequency power to high-frequency power of heart rate variability and heart rate-respiratory cross-consistency.

[0045] The fatigue driving comprehensive evaluation module is used to calculate the fatigue driving comprehensive evaluation index using a weighted method based on the driving fatigue evaluation indicators.

[0046] According to specific embodiments provided by the present invention, the present invention discloses the following technical effects:

[0047] Compared to existing technologies that rely on indirect methods such as assessing driver fatigue based on external behavioral characteristics or vehicle driving status, or on individual physiological signals, which suffer from low accuracy, this invention comprehensively considers both psychological and physiological fatigue. It assesses driver fatigue from three dimensions: physical load, alertness, and cognitive load. It integrates indicators such as the root mean square value of electromyography (EMG) signals, the integral value of EMG signals, the mean heart rate, the root mean square value of the difference between adjacent neural intervals throughout the journey, the ratio of low-frequency to high-frequency power in heart rate variability, and the consistency of heart rate and respiration. Using a weighted method, this approach effectively improves the accuracy and objectivity of driver fatigue assessment. Attached Figure Description

[0048] To more clearly illustrate the technical solutions in the embodiments of the present invention 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 the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0049] Figure 1 A flowchart of an intelligent driving fatigue evaluation method provided in an embodiment of the present invention;

[0050] Figure 2 A diagram illustrating the evaluation index system provided in this embodiment of the invention;

[0051] Figure 3 The diagram illustrates an intelligent driving fatigue evaluation model based on driver physiological and psychological responses, provided as an embodiment of the present invention. Detailed Implementation

[0052] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0053] The purpose of this invention is to provide a method and system for evaluating driver fatigue.

[0054] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0055] Example 1

[0056] See Figure 1 and Figure 2 This embodiment provides a method for evaluating fatigue in intelligent driving, the method comprising:

[0057] Step 100: Acquire the driver's physiological data in the driving scenario; the physiological data includes electromyography data, heart rate data, and respiratory rate data;

[0058] Step 200: Calculate driving fatigue evaluation indicators based on the physiological data; the driving fatigue evaluation indicators include physiological driving fatigue evaluation indicators and psychological driving fatigue evaluation indicators; wherein, the physiological driving fatigue evaluation indicators include physical load indicators; the physical load indicators include the root mean square value of electromyography (EMG) signals and the integral value of EMG signals; the psychological driving fatigue evaluation indicators include alertness indicators and cognitive load indicators; the alertness indicators include the mean heart rate, the root mean square value of the difference between adjacent NN intervals throughout the journey, and the ratio of low-frequency power to high-frequency power of heart rate variability, where the NN interval represents the interval between two adjacent heartbeat fluctuations; the cognitive load indicators include the ratio of low-frequency power to high-frequency power of heart rate variability and heart rate-respiratory cross-consistency;

[0059] Step 300: Calculate the comprehensive fatigue driving evaluation index using the weighting method based on the driving fatigue evaluation indicators.

[0060] In this embodiment, the intelligent cockpit fatigue driving evaluation index is designed based on physiological and psychological factors. It characterizes the driver's overall fatigue status from three dimensions: physical load, alertness, and cognitive load. Through feature fusion of multiple indicators, an intelligent cockpit fatigue driving evaluation model based on the driver's physiological and psychological reactions is constructed.

[0061] In the selection of various evaluation indicators, the physiological data indicators of the driving process are compared with the indicators under natural conditions, and the selection is carried out according to the principles of significance, specificity and effectiveness.

[0062] Figure 2 In the diagram, the symbols "+" and "-" below each indicator indicate positive and negative correlations, respectively.

[0063] In some embodiments, the formula for calculating the integral value of electromyography (iEMG) in step 200 is as follows:

[0064]

[0065] Where Fs is the sampling frequency, EMG iThis represents the amplitude of the i-th electromyographic signal, and N is the number of electromyographic signal amplitude data.

[0066] In some embodiments, in step 200, the formula for calculating the root mean square (RMS) value of the electromyographic signal is as follows:

[0067]

[0068] Physical exertion was measured using root mean square (RMS) and integrated electromyography (IEMGM) values. Electromyography (EMG) is the temporal and spatial superposition of action potentials of motor units within numerous muscle fibers. Surface electromyography (sEMG) is the combined effect of superficial muscle EMG and electrical activity on nerve trunks on the skin surface, reflecting neuromuscular activity to a certain extent. As a non-invasive, real-time measurement method, surface electromyography can objectively reflect muscle activity levels and functional states, and it also has significant practical value in assessing driving fatigue. RMS and iEMG can effectively characterize the dynamic changes in physiological fatigue.

[0069] The root mean square (RMS) value of electromyography (EMG) signal refers to the effective value of muscle discharge. Its magnitude varies depending on the change in EMG amplitude, reflecting the average level of muscle discharge over a certain period of time. From the initial state to the fatigue state, the amplitude of the surface EMG signal increases. That is, as fatigue increases, RMS increases and is positively correlated with physical load.

[0070] The integrated electromyography (iEMG) value refers to the total electrical discharge of motor units involved in muscle activity over a certain period of time. Generally, the larger the value, the more severe the fatigue. It is an important indicator for evaluating muscle fatigue and is positively correlated with physical load.

[0071] In some embodiments, the ratio of low-frequency power to high-frequency power of heart rate variability (LF / HF) is calculated using the following formula:

[0072] LF = Sum(FFT(RR)) | 0.04～0.15Hz (3);

[0073] HF = Sum(FFT(RR)) | 0.15～0.4Hz (4);

[0074] Where LF is the low-frequency power of heart rate variability, HF is the high-frequency power of heart rate variability, Sum() represents the summation operation, FFT() represents the fast Fourier transform operation, RR is the heartbeat interval, and the RR interval represents the time interval between two adjacent heartbeats (R waves), usually in milliseconds (ms).

[0075] FFT(RR) refers to analyzing heart rate interval data using Fast Fourier Transform to convert it into a frequency spectrum in order to study the frequency domain characteristics of heart rate variability.

[0076] Sum()|0.04~0.15 indicates that the frequency range of LF is not fixed, and usually includes the frequency range between 0.04Hz and 0.15Hz.

[0077] In some embodiments, in step 200, the formula for calculating the mean heart rate (MeanHR) is:

[0078]

[0079] Among them, HR i Let N be the heart rate at the i-th minute, and N be the number of heart rate data sets, with each minute's heart rate data set constituting one set.

[0080] In some embodiments, in step 200, the root mean square value (RMSSD) of the difference between adjacent NN intervals throughout the entire process is calculated using the following formula:

[0081]

[0082] Among them, RR i RR i+1 These are the i-th and (i+1)-th heartbeat intervals, respectively; N is the number of heartbeat interval data.

[0083] Alertness was measured using mean heart rate, RMSSD, and LF / HF. Since the heart's automatic activity is controlled by the sympathetic and parasympathetic nervous systems, heart rate variability (HRV) contains rich information on the body's cardiovascular self-balancing regulation. Characteristic indicators that can represent the physiological information of the tested individual are extracted from this regulatory information. Changes in heart rate are closely related to physiological fatigue and arousal levels. Heart rate (HR) and heart rate variability (HRV) are currently the two most commonly used indicators in fatigue detection research.

[0084] Mean heart rate (MeanHR) refers to the number of heartbeats per minute and is positively correlated with alertness.

[0085] RMSSD is a time-domain indicator of heart rate variability (HRV), which refers to the root mean square value of the difference between adjacent NN intervals throughout the entire process. It reflects the fast-changing component of HRV, characterizes parasympathetic nerve activity, and is negatively correlated with alertness.

[0086] LF / HF is a frequency domain index of heart rate variability (HRV), which refers to the ratio of low-frequency power to high-frequency power of HRV. It is an indicator of the balance between the sympathetic and vagus nerves, and is positively correlated with sympathetic nerve activity, positively correlated with alertness, and negatively correlated with cognitive load.

[0087] In some embodiments, the formula for calculating the heart rate-respiration cross-consistency coefficient (CPC) in step 200 is as follows:

[0088]

[0089] Where HR is heart rate, Resp is respiratory rate, Cross() calculates the cross-correlation between signals, and Coherence() calculates the coherence between signals.

[0090] Cognitive load was measured using LF / HF and heart rate-respiratory cross-congruence (CPC). Heart rate-respiratory cross-congruence characterizes the driver's relaxation state during driving; the more relaxed the driver, the lower the psychological load.

[0091] The heart rate-respiration cross-consistency coefficient (CPC) refers to the consistency of heart rate and respiratory rate rhythms in a conscious state, which characterizes the driver's psychological load. A higher cardiopulmonary cross-consistency value indicates a higher level of user attention, meaning that the user is using more attentional resources and is more focused on a certain thing, and that their cognitive load is higher. This indicator is positively correlated with cognitive load.

[0092] In some embodiments, step 300 calculates the comprehensive fatigue driving evaluation index using a weighted method based on the driving fatigue evaluation index, specifically including:

[0093] The fatigue driving comprehensive evaluation index A is calculated using the following formula:

[0094] A=μ1A1+μ2A2+μ3A3+μ4A4+μ5A5+μ6A6, (8);

[0095] Where A1 is the root mean square value of the electromyography (EMG) signal, A2 is the integral value of the EMG signal, A3 is the root mean square value of the difference between adjacent NN intervals throughout the entire process, A4 is the ratio of low-frequency power to high-frequency power of heart rate variability, A5 is the mean heart rate, and A6 is the heart rate-respiration cross-consistency coefficient; μ1, μ 12 ,…,μ6 are the weight coefficients of A1, A2,…,A6 respectively.

[0096] In some embodiments, the formula for calculating the weighting coefficients μ1-μ6 is as follows:

[0097]

[0098] Where, μ i Let A be the i-th driver fatigue evaluation index. i Weighting coefficients; W i The weight of the i-th driver fatigue evaluation index determined by the analytic hierarchy process; w i is the weight of the i-th driving fatigue evaluation index determined by the entropy weight method; N is the total number of driving fatigue evaluation indicators.

[0099] This embodiment uses the AHP entropy method for comprehensive weighting (combining the analytic hierarchy process weighting with the entropy method weighting) to evaluate the fatigue driving safety risk of intelligent cockpits based on physiological and psychological factors. The higher the index value, the higher the degree of physiological and psychological fatigue of the driver and the lower the safety level.

[0100] In some embodiments, the process of determining the weight coefficient of the i-th driving fatigue evaluation index using the analytic hierarchy process is as follows:

[0101] The relative importance of each driver fatigue evaluation indicator was scored using the priority relation quantitative scaling method, and a judgment matrix was obtained.

[0102] The judgment matrix is ​​normalized column-wise, and the normalized judgment matrix is ​​then summed row-wise to obtain the sum vector M. i , where i represents the i-th driver fatigue evaluation index;

[0103] The weight coefficient W of the i-th driving fatigue evaluation index is calculated according to the following formula. i :

[0104]

[0105] Where N represents the total number of driver fatigue evaluation indicators.

[0106] In some embodiments, the process of determining the weight coefficient of the i-th driving fatigue evaluation index using the analytic hierarchy process (AHP) further includes:

[0107] The relative importance of each driver fatigue evaluation indicator was scored using a priority relation quantitative scaling method, with a scale of 0.1 to 0.9 as shown in Table 2. An expert scoring method was used to assign importance values ​​to six indicators: F1 (root mean square electromyography value), F2 (integrated electromyography value), F3 (RMSSD), F4 (LF / HF), F5 (mean heart rate), and F6 (heart rate-respiratory cross-consistency coefficient). The resulting matrix, Table 1, shows the judgment matrix A.

[0108] Table 1 Judgment Matrix A

[0109] <![CDATA[F1]]> <![CDATA[F2]]> <![CDATA[F3]]> <![CDATA[F4]]> <![CDATA[F5]]> <![CDATA[F6]]> <![CDATA[F1]]> <![CDATA[A 11 ]]> <![CDATA[A 12 ]]> <![CDATA[A 13 ]]> <![CDATA[A 14 ]]> <![CDATA[A 15 ]]> <![CDATA[A 16 ]]> <![CDATA[F2]]> <![CDATA[A 21 ]]> <![CDATA[A 22 ]]> <![CDATA[A 23 ]]> <![CDATA[A 24 ]]> <![CDATA[A 25 ]]> <![CDATA[A 26 ]]> <![CDATA[F3]]> <![CDATA[A 31 ]]> <![CDATA[A 32 ]]> <![CDATA[A 33 ]]> <![CDATA[A 34 ]]> <![CDATA[A 35 ]]> <![CDATA[A 36 ]]> <![CDATA[F4]]> <![CDATA[A 41 ]]> <![CDATA[A 42 ]]> <![CDATA[A 43 ]]> <![CDATA[A 44 ]]> <![CDATA[A 45 ]]> <![CDATA[A 46 ]]> <![CDATA[F5]]> <![CDATA[A 51 ]]> <![CDATA[A 52 ]]> <![CDATA[A 53 ]]> <![CDATA[A 54 ]]> <![CDATA[A 55 ]]> <![CDATA[A 56 ]]> <![CDATA[F6]]> <![CDATA[A 61 ]]> <![CDATA[A 62 ]]> <![CDATA[A 63 ]]> <![CDATA[A 64 ]]> <![CDATA[A 65 ]]> <![CDATA[A 66 ]]>

[0110] Table 2: Judgment Matrix Scale and Meaning

[0111]

[0112]

[0113] After obtaining the judgment matrix A using the above method, normalize matrix A by column and then perform row averaging to obtain the weight coefficients.

[0114] For example, matrix A = [a ij ] N×N Where N is the number of driving fatigue evaluation indicators (number of elements, the same below). Matrix A is normalized using the following formula:

[0115]

[0116] Add the normalized judgment matrices row by row to obtain the sum vector M. i As shown in the following formula,

[0117]

[0118] Then, the row average is performed according to formula (10) to obtain the weighting coefficient W. i .

[0119] In some embodiments, before performing the step of normalizing the judgment matrix column-wise, the method further includes:

[0120] If the number of driving fatigue evaluation indicators is greater than 2, then the judgment matrix is ​​subjected to a consistency check. If the judgment matrix fails the consistency check, then the judgment matrix is ​​modified and the consistency check is performed again until the judgment matrix passes the consistency check.

[0121] If the number of driving fatigue evaluation indicators is less than or equal to 2, then proceed to the step of normalizing the judgment matrix column by column.

[0122] The purpose of performing a consistency check on the judgment matrix is ​​to ensure logical consistency.

[0123] In some embodiments, performing a consistency check on the judgment matrix further includes:

[0124] The largest eigenvalue γ of the judgment matrix A is calculated using the square root method. max As shown in the following formula:

[0125]

[0126] Among them, A*W i To determine the eigenvalues ​​of a matrix, W i Let be the weight coefficient of the i-th driving fatigue evaluation index, N be the number of driving fatigue evaluation indexes, and n be the number of dimensions (with the same meaning as n in Table 3), N = n.

[0127] The consistency index CI is calculated using the following formula, and the average random consistency index CR is obtained from Table 3.

[0128]

[0129] Table 3. Values ​​of the average random consistency index CR

[0130] n 3 4 5 6 7 8 9 10 11 CR 0.58 0.9 1.12 1.24 1.32 1.41 1.45 1.49 1.51

[0131] If the following inequality is satisfied, the consistency check passes; otherwise, the judgment matrix needs to be modified and the consistency check needs to be performed again until it passes.

[0132]

[0133] In some embodiments, modifying the judgment matrix includes:

[0134] 1. Modify the judgment matrix data. For small error accumulation in multiple data points of the judgment matrix: minimize the modification of expert data and fine-tune each element of the matrix; for judgment errors in a certain data point of the judgment matrix: find the element with the highest deviation in the judgment matrix process and correct it.

[0135] 2. Reduce the order of the judgment matrix.

[0136] In some embodiments, the process of determining the weight coefficient of the i-th driving fatigue evaluation index using the entropy weight method is as follows:

[0137] The raw data of each driver fatigue evaluation indicator are standardized to obtain unit data for each indicator. The information entropy and information utility values ​​of each indicator are then calculated using the following formulas:

[0138]

[0139] d i =1-e i (17);

[0140] Among them, e i Let d be the information entropy value of the i-th driving fatigue evaluation index. i Let m be the information utility value of the i-th driver fatigue evaluation index. * Y represents the number of raw data points for a single driver fatigue evaluation indicator. ti This represents the t-th unit of data for the i-th driving fatigue evaluation index.

[0141] The weight coefficient w of the i-th driving fatigue evaluation index determined by the entropy weight method is calculated using the following formula. i :

[0142]

[0143] The basic idea of ​​the entropy weight method is: if the value of a certain indicator changes more, the degree of variation of the indicator value is greater, the more information it provides, the smaller the information entropy value, and the greater the weight; conversely, the smaller the entropy value, the smaller the weight.

[0144] In some embodiments, the t-th unit data Y of the i-th driving fatigue evaluation index ti Calculated as follows:

[0145] First, the original driver fatigue evaluation index data is standardized as follows:

[0146]

[0147] Where x ti Let x be the raw value of the t-th unit of data for the i-th driving fatigue evaluation index. imax Let x be the maximum raw value of the i-th driving fatigue evaluation index per unit. imin Y represents the smallest unit of raw data for the i-th driver fatigue evaluation index. ti It is the t-th unit data of the i-th driver fatigue evaluation index after dimensionless processing, and its standardization is defined as follows:

[0148]

[0149] In the formula, m * This represents the number of raw data points for a single driver fatigue evaluation indicator.

[0150] See Figure 3 In some embodiments, a method for evaluating fatigue in intelligent driving further includes:

[0151] Design a smart cockpit driving simulation experiment based on physiological and psychological factors.

[0152] Construct an intelligent cockpit fatigue driving evaluation index system based on the driver's physiological and psychological reactions.

[0153] Establish an intelligent driving fatigue evaluation method based on the driver's physiological and psychological responses.

[0154] In some embodiments, the physiological and psychological-based intelligent cockpit driving simulation experiment includes:

[0155] (1) The experimental sample consisted of 40 drivers with more than 2 years of driving experience and a driver's license level of C2 or above. The average age distribution was 18-25 years old, 25-35 years old, 35-45 years old, and 45 years old and above, with a male-to-female ratio of 1:1.

[0156] (2) The experimental equipment consists of two parts: a driving simulator and a data acquisition system. The data acquisition system includes electromyography (EMG) and electrocardiography (ECG) devices. The EMG measurement sites are the left and right trapezius muscles and calf muscles, and the ECG measurement sites are the left and right sides of the sternum.

[0157] (3) Drivers participated in two 90-minute driving simulation experiments. For the first experiment, drivers had 8 hours of sleep the night before, and for the second experiment, they had 4 hours of sleep the night before, from 3:00 AM to 7:00 AM. Volunteers wore wristbands to monitor their sleep time before going to sleep. Both experiments started at 9:00 AM. Chewing gum and making or receiving phone calls were prohibited during the experiments, and the experimental environment was kept quiet to avoid external stimuli affecting the drivers' performance.

[0158] (4) The road environment in the simulator is a closed circular composite road, including urban, rural, highway and mountain roads, with a small number of vehicles on the road. Drivers are free to drive as long as they do not exceed the speed limit and do not cause accidents. The experiment is conducted indoors, with good air circulation, a constant room temperature of 23°C and stable and sufficient light.

[0159] Example 2

[0160] This embodiment provides an intelligent driving fatigue assessment system, including:

[0161] The physiological data acquisition module is used to acquire the driver's physiological data, including electromyography data, heart rate data, and respiratory rate data.

[0162] A driver fatigue evaluation index calculation module is used to calculate driver fatigue evaluation indices based on the physiological data. These indices include physiological driver fatigue evaluation indices and psychological driver fatigue evaluation indices. The physiological driver fatigue evaluation indices include physical load indices, which include the root mean square value of electromyography (EMG) signals and the integral value of EMG signals. The psychological driver fatigue evaluation indices include alertness indices and cognitive load indices. The alertness indices include the mean heart rate, the root mean square value of the difference between adjacent NN intervals throughout the journey, and the ratio of low-frequency power to high-frequency power of heart rate variability, where the NN interval represents the interval between two adjacent heartbeats. The cognitive load indices include the ratio of low-frequency power to high-frequency power of heart rate variability and heart rate-respiratory cross-consistency.

[0163] The fatigue driving comprehensive evaluation module is used to calculate the fatigue driving comprehensive evaluation index using a weighted method based on the driving fatigue evaluation indicators.

[0164] In summary, the present invention has the following advantages:

[0165] This invention proposes an intelligent driving fatigue evaluation method based on driver physiological and psychological responses, achieving the simultaneous detection of psychological and physiological fatigue during driving. A refined evaluation index is selected to construct the evaluation system, and the analytic hierarchy process (AHP) and entropy weight evaluation method are employed to construct the evaluation method. This largely eliminates the reliance on index data and provides valuable reference for fatigue driving safety evaluation in my country's intelligent cockpit automotive industry.

[0166] This invention is based on an intelligent driving fatigue evaluation system that uses the driver's physiological and psychological reactions to characterize the driver's physiological and psychological behavioral characteristics from three dimensions: physical load, alertness, and cognitive load. It constructs an intelligent cockpit fatigue driving evaluation index system, which overcomes the shortcomings of low accuracy in identifying single parameters.

[0167] The technical solution provided by this invention enables accurate evaluation of driver fatigue in intelligent cockpits, which is beneficial for promotion and application. It can significantly reduce the incidence of serious traffic accidents caused by driver fatigue in intelligent cockpits, and also provides a reference for driving safety status monitoring and intelligent cockpit safety design.

[0168] This invention addresses fatigue driving behavior in intelligent cockpits by employing a combination of subjective and objective methods. It uses the analytic hierarchy process (AHP) and entropy weight method to determine the weights of indicators, thus possessing both subjective rationality and objective accuracy.

[0169] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the systems disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the descriptions are relatively simple; relevant parts can be referred to the method section.

[0170] This document uses specific examples to illustrate the principles and implementation methods of the present invention. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of the present invention. Furthermore, those skilled in the art will recognize that, based on the ideas of the present invention, 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 the present invention.

Claims

1. A method for evaluating intelligent driving fatigue, characterized by, The method comprises: acquiring physiological data of a driver in a driving scene; the physiological data comprises electromyography data, heart rate data and respiration rate data; calculating a driving fatigue evaluation index according to the physiological data; the driving fatigue evaluation index comprises a physiological driving fatigue evaluation index and a psychological driving fatigue evaluation index; the physiological driving fatigue evaluation index comprises a physical load index; the physical load index comprises an electromyography signal root mean square value and an electromyography signal integral value; the psychological driving fatigue evaluation index comprises an alertness index and a cognitive load index; the alertness index comprises a mean heart rate, a root mean square value of a difference between adjacent NN intervals and a ratio of a low frequency power to a high frequency power of heart rate variability; the NN interval represents an interval between adjacent two heartbeats; the cognitive load index comprises the ratio of the low frequency power to the high frequency power of the heart rate variability and a heart rate respiration cross consistency; calculating a comprehensive fatigue driving evaluation index by using a weight method according to the driving fatigue evaluation index; specifically comprising: calculating the comprehensive fatigue driving evaluation index A by using the following formula: A = μ1A1 + μ2A2 + μ3A3 + μ4A4 + μ5A5 + μ6A6, Where A1 is the root mean square value of the electromyography (EMG) signal, A2 is the integral value of the EMG signal, A3 is the root mean square value of the difference between adjacent NN intervals throughout the entire process, A4 is the ratio of low-frequency power to high-frequency power of heart rate variability, A5 is the mean heart rate, and A6 is the heart rate-respiration cross-consistency coefficient; μ1, μ 12 μ1, ..., μ6 are the weight coefficients of A1, A2, ..., A6, respectively; a calculation formula of the weight coefficients μ1-μ6 is: wherein μ i is the weight coefficient of the ith driving fatigue evaluation index A i ; W i is the weight of the ith driving fatigue evaluation index determined by the analytic hierarchy process; w i is the weight of the ith driving fatigue evaluation index determined by the entropy weight method; and N is the total number of driving fatigue evaluation indexes. 2.The intelligent driving fatigue evaluation method according to claim 1, characterized in that, a calculation formula of the electromyography signal integral value iEMG is: where Fs is the sampling frequency, EMG i represents the i-th EMG signal amplitude, and N is the number of EMG signal amplitude data. 3.The intelligent driving fatigue evaluation method according to claim 1, characterized in that, calculating the ratio LF / HF of the low frequency power to the high frequency power of the heart rate variability by using the following formula: LF = Sum e (FFT(RR)) |2 0.04～0.15Hz , HF = Sum∈(FFT(RR)) |2 0.15～0.4Hz , wherein, LF is the low frequency power of the heart rate variability, HF is the high frequency power of the heart rate variability, Sum() represents a summation operation, FFT() represents a fast Fourier transform operation, and RR is a heartbeat interval. 4.The intelligent driving fatigue evaluation method according to claim 1, characterized in that, a calculation formula of the heart rate respiration cross consistency coefficient CPC is: wherein, HR is a heart rate, Resp is a respiration rate, Cross() represents a cross correlation between signals, and Coherence() represents a coherence between signals. 5.The intelligent driving fatigue evaluation method according to claim 1, characterized in that, a process of determining the weight coefficient of the i-th driving fatigue evaluation index by using an analytic hierarchy process is: scoring the relative importance of each driving fatigue evaluation index by using a priority relationship quantity scale method to obtain a judgment matrix; The judgment matrix is normalized by column, and the normalized judgment matrix is added by row to obtain a sum vector M i , i represents the i-th driving fatigue evaluation index The weight coefficient W of the ith driving fatigue evaluation index is calculated according to the following formula i : 6.The intelligent driving fatigue evaluation method according to claim 5, characterized in that, before performing the step of normalizing the judgment matrix by column, further comprising: if the number of the driving fatigue evaluation indexes is greater than 2, performing a consistency check on the judgment matrix; if the judgment matrix does not pass the consistency check, modifying the judgment matrix and re-performing the consistency check until the judgment matrix passes the consistency check; if the number of the driving fatigue evaluation indexes is less than or equal to 2, jumping to the step of normalizing the judgment matrix by column. 7.The intelligent driving fatigue evaluation method according to claim 1, characterized in that, a process of determining the weight coefficient of the i-th driving fatigue evaluation index by using an entropy weight method is: standardizing the original data of each driving fatigue evaluation index to obtain unit data of each driving fatigue evaluation index, and calculating an information entropy value and an information utility value of each driving fatigue evaluation index according to the following formula: d i = 1 - e i , Wherein, e i is the information entropy value of the ith driving fatigue evaluation index, d i is the information utility value of the ith driving fatigue evaluation index, m * is the number of original data of a single driving fatigue evaluation index, Y ti is the tth unit data of the ith driving fatigue evaluation index; The weight coefficient w of the ith driving fatigue evaluation index determined by the entropy weight method is calculated according to the following formula i :

8. An intelligent driving fatigue evaluation system, characterized by comprising: comprising: a physiological data acquisition module configured to acquire physiological data of a driver; the physiological data comprises electromyography data, heart rate data and respiration rate data; a driving fatigue evaluation index calculation module configured to calculate a driving fatigue evaluation index according to the physiological data; The driving fatigue evaluation index includes a physiological driving fatigue evaluation index and a psychological driving fatigue evaluation index; the physiological driving fatigue evaluation index includes a physical load index; the physical load index includes an electromyography signal root mean square value and an electromyography signal integral value; the psychological driving fatigue evaluation index includes an alertness index and a cognitive load index; the alertness index includes a heart rate mean value, a root mean square value of a difference between adjacent NN intervals, and a ratio of a low frequency power to a high frequency power of heart rate variability, the NN interval representing an interval between adjacent two heartbeat fluctuations; the cognitive load index includes the ratio of the low frequency power to the high frequency power of the heart rate variability and a heart rate respiration cross consistency; The fatigue driving comprehensive evaluation module is configured to calculate a fatigue driving comprehensive evaluation index according to the driving fatigue evaluation index by using a weight method; specifically including: The fatigue driving comprehensive evaluation index A is calculated by using the following formula: A = μ1A1 + μ2A2 + μ3A3 + μ4A4 + μ5A5 + μ6A6, Where A1 is the root mean square value of the electromyography (EMG) signal, A2 is the integral value of the EMG signal, A3 is the root mean square value of the difference between adjacent NN intervals throughout the entire process, A4 is the ratio of low-frequency power to high-frequency power of heart rate variability, A5 is the mean heart rate, and A6 is the heart rate-respiration cross-consistency coefficient; μ1, μ 12 μ1, ..., μ6 are the weight coefficients of A1, A2, ..., A6, respectively; The calculation formula of the weight coefficients μ1-μ6 is: wherein μ i is the weight coefficient of the ith driving fatigue evaluation index A i ; W i is the weight of the ith driving fatigue evaluation index determined by the analytic hierarchy process; w i is the weight of the ith driving fatigue evaluation index determined by the entropy weight method; and N is the total number of driving fatigue evaluation indexes.

Images