CN-121973792-A - Intelligent intervention method and system for driving fatigue

CN121973792ACN 121973792 ACN121973792 ACN 121973792ACN-121973792-A

Abstract

The invention discloses a method and a system for intelligent intervention of driving fatigue, which relate to the technical field of intelligent cabin control and comprise the following steps of continuously collecting physiological state data of a driver and cabin environment data in the running process of a vehicle; the method comprises the steps of performing multi-mode preprocessing, feature extraction and fusion analysis on physiological state data to obtain NCI, calculating a first derivative of the NCI, constructing a cabin thermal energy index based on cabin environment data, judging whether a conventional temperature control strategy is triggered or not based on the NCI and the first derivative thereof and the cabin thermal energy index, calculating and generating a control instruction when the conventional temperature control strategy is triggered, executing a spatial asymmetric thermal strategy when the NCI is not obviously improved in a preset observation window after the conventional temperature control strategy is executed, re-collecting the physiological state data and the cabin environment data after any strategy is executed, updating the NCI and the first derivative thereof, evaluating the executing effect of the strategy and feeding back.

Inventors

ZHAO ZHONGZE

Assignees

上海凌芯体育科技有限公司

Dates

Publication Date: 20260505
Application Date: 20260323

Claims (9)

1. A method for intelligently intervening driving fatigue is characterized by comprising the following steps: Continuously collecting physiological state data of a driver and cabin environment data in the running process of the vehicle; Preprocessing, feature extraction and fusion analysis are carried out on the physiological state data to obtain a real-time neural-cognitive impairment index NCI; Constructing cabin thermal energy indexes based on cabin environment data; judging whether to trigger a conventional temperature control strategy or not based on NCI and a first derivative thereof and cabin thermal energy index, and calculating and generating a control instruction when the conventional temperature control strategy is triggered; In a preset observation window after the conventional temperature control strategy is executed, executing a space asymmetric thermal strategy when NCI is not obviously improved; after any strategy is executed, physiological state data and cabin environment data are collected again, NCI and first derivative thereof are updated, and strategy execution effect is evaluated and feedback is carried out.
2. The method for intelligent intervention of driving fatigue according to claim 1, wherein the continuously collecting the driver physiological state data and the cabin environment data specifically comprises the following steps: the method comprises the steps that an original time sequence of physiological state data and cabin environment data is collected cooperatively through a vehicle-mounted sensor and a wearable device; The physiological state data comprises Heart Rate Variability (HRV) signals and myoelectricity (EMG) signals; Preprocessing and extracting features of the collected HRV and EMG physiological data to obtain HRV feature parameters and EMG feature parameters; inputting the HRV characteristic parameters and the EMG characteristic parameters into a pre-trained machine learning model for fusion analysis, and calculating to obtain a neural-cognitive impairment index NCI reflecting the reduction degree of the neural and cognitive functions of a driver; The HRV characteristic parameters comprise time domain characteristic parameters RMSSD and SDNN, frequency domain characteristic parameters LF/HF ratio and nonlinear characteristic parameter sample entropy, wherein the EMG characteristic parameters comprise root mean square value RMS of electromyographic signal amplitude, total power or spectrum energy in a preset frequency band and muscle fatigue index determined according to high-frequency component change of the electromyographic power spectrum; and (5) unifying a time reference, and completing time stamp alignment of the physiological state data and the cabin environment data.
3. The method for intelligent intervention of driving fatigue according to claim 2, wherein the NCI first derivative is calculated by a smooth difference method; Based on the historical NCI sequence, predicting an NCI track in future time length by adopting an exponential smoothing prediction method; The NCI preset threshold value comprises a first demarcation value and a second demarcation value; the first demarcation value is used for dividing the safe area from the light fatigue area, and the second demarcation value is used for dividing the light fatigue area from the heavy fatigue area.
4. The method for intelligently intervening in the driving fatigue according to claim 3, wherein the cabin environment data is an environment parameter set for representing the hot environment and the surrounding state of the cabin; The cabin thermal kinetic energy index is constructed through a parameterized approximate model, and a model formula is H (t) =k1.x1+k2.x2+k3.x3+k4.x4; Wherein X1 is the normalized cabin temperature change rate, X2 is the difference between the normalized air conditioner set temperature and the actual cabin temperature, X3 is the normalized air conditioner operating power, X4 is the normalized temperature difference between the inside and outside of the vehicle, and k1, k2, k3 and k4 are dimensionless weight coefficients subjected to history data calibration; And predicting the change trend of the cabin thermal energy index in the future time length by adopting a linear trend prediction method.
5. The method for intelligent intervention of driving fatigue according to claim 4, wherein the determining whether to trigger the conventional temperature control strategy specifically comprises: when the NCI is in a safety zone defined by a first demarcation value, the first derivative of the NCI is continuously negative, the absolute value of the NCI presents an increasing situation, and the cabin thermal energy index is in an increasing situation, a conventional temperature control strategy is triggered; the calculating and generating the control instruction specifically comprises the following steps: Based on the absolute value of the NCI first derivative and the increment rate of the cabin thermal energy index, calculating to obtain an environment regulation parameter corresponding to a conventional temperature control strategy through a preset mapping function; the environment regulation parameters comprise air conditioner air quantity regulation quantity, air outlet torsion angle and air circulation rate regulation quantity; The control instructions are generated based on the environment adjustment parameters, the control instructions are issued, and a conventional temperature control strategy is executed based on the control instructions.
6. The method for intelligent intervention of driving fatigue according to claim 5, wherein the preset mapping function is a mapping relation model constructed based on the corresponding relation of NCI first derivative absolute value, cabin thermal kinetic energy index increment rate and environment adjustment parameter, and is obtained by executing sample training calibration through a history strategy, and is used for realizing the corresponding relation of input parameter and environment adjustment parameter.
7. The method for intelligent intervention of driving fatigue according to claim 6, wherein the preset observation window is used for judging the execution effect of the conventional temperature control strategy, and the NCI index is determined according to a preset amplitude judgment rule without significant improvement; And the first boundary value and the second boundary value are used as judging reference bases for judging whether NCI is improved remarkably, and whether NCI indexes reach a remarkable improvement standard is determined jointly by combining a preset amplitude judging rule.
8. The method for intelligent intervention of driving fatigue according to claim 7, wherein the executing the spatially asymmetric thermal strategy comprises: Dividing the cabin into a plurality of temperature control partitions, wherein the temperature control partitions comprise left and right side partitions or head and foot partitions; based on the current NCI and the improvement amplitude of NCI and the cabin thermal kinetic energy index, calculating environmental regulation parameters corresponding to each temperature control partition through a preset partition parameter mapping model; The environment regulation parameters comprise target temperature offset of each partition, wind speed regulation of each partition air outlet, torsion angle of each partition air outlet and partition temperature control switching period; the preset partition parameter mapping model is obtained by performing sample training calibration by taking the improvement amplitude of the current NCI and NCI, the cabin thermal kinetic energy index and the partition sensitivity parameter as input variables and taking the environment adjustment parameters of each temperature control partition as output variables and combining a history space asymmetric thermal strategy, and the environment adjustment parameters corresponding to each temperature control partition are obtained through calculation; the partition sensitivity parameter refers to a preset quantization parameter used for representing the adjustment sensitivity degree of a driver to different temperature control partition environments of the cabin; The improvement amplitude of the NCI is determined by combining the first demarcation value, the second demarcation value and a preset amplitude judgment rule for the difference value of the NCI before and after the conventional temperature control strategy is executed.
9. A system for intelligent intervention of driving fatigue is characterized by being applied to the method for intelligent intervention of driving fatigue according to any one of claims 1-8.

Description

Intelligent intervention method and system for driving fatigue Technical Field The invention relates to the technical field of intelligent cabin control, in particular to a method and a system for intelligent intervention of driving fatigue. Background With the popularization of automobile intellectualization, a driver status monitoring system (DMS) has become a key technology for improving driving safety. Driving fatigue is one of the key factors causing traffic accidents, and human body has physiological characteristics of slow response, reduced cognitive ability and the like in a fatigue state. Traditional fatigue reminding mostly depends on behavior characteristics such as wink, yawning and the like captured by a visual sensor, or compulsively reminding in modes such as vibration seats, voice warning and the like. Meanwhile, the cabin thermal environment has obvious influence on the arousal degree of people, and the physiological state of a driver is improved by adjusting environmental parameters such as temperature, air quantity and the like, so that the cabin thermal environment becomes a hot spot direction for research in the field of intelligent cabins. These studies provide a theoretical basis for achieving "soft intervention" of driving fatigue through physical environment interactions. The existing driving fatigue intervention technology still has a plurality of limitations in practical application, namely, the existing cabin temperature control strategy is mainly unified regulation of the whole cabin, the influence of cabin thermal inertia and environmental dynamic energy consumption on the intervention effect is not fully considered, and the regulation mode is mechanized. In addition, existing systems lack effective closed loop feedback and advanced intervention mechanisms, and when conventional temperature control approaches fail due to driver physiological adaptability, the systems fail to provide a more targeted stimulation scheme. Disclosure of Invention The invention aims to provide a method and a system for intelligently intervening driving fatigue, which are used for solving the problems in the prior art. In order to achieve the purpose, the invention provides the following technical scheme that the intelligent intervention method for driving fatigue comprises the following steps: Continuously collecting physiological state data of a driver and cabin environment data in the running process of the vehicle; Performing multi-mode pretreatment, feature extraction and fusion analysis on the physiological state data to obtain a real-time neural-cognitive impairment index NCI, and calculating a NCI first derivative; Constructing cabin thermal energy indexes based on cabin environment data; judging whether to trigger a conventional temperature control strategy or not based on NCI and a first derivative thereof and cabin thermal energy index, and calculating and generating a control instruction when the conventional temperature control strategy is triggered; In a preset observation window after the conventional temperature control strategy is executed, executing a space asymmetric thermal strategy when NCI is not obviously improved; after any strategy is executed, physiological state data and cabin environment data are collected again, NCI and first derivative thereof are updated, and strategy execution effect is evaluated and feedback is carried out. The continuously collecting driver physiological state data and cabin environment data specifically comprises the following steps: the vehicle-mounted sensor can be an infrared heart rate sensor and a steering wheel myoelectric sensor in the cabin, the wearable device can be an intelligent bracelet and an intelligent helmet, and the collaborative acquisition can avoid the loss of data or overlarge error acquired by a single device; The physiological state data comprises Heart Rate Variability (HRV) signals and myoelectricity (EMG) signals; Preprocessing and extracting features of the collected HRV and EMG physiological data to obtain HRV feature parameters and EMG feature parameters; Inputting the HRV characteristic parameters and the EMG characteristic parameters into a pre-trained machine learning model for fusion analysis, and calculating to obtain a driving nerve-cognition damage index NCI reflecting the reduction degree of the driver nerve and cognition function; The HRV characteristic parameters comprise a time domain characteristic parameter RMSSD (root mean square for calculating the RR interval difference value of continuous heartbeat) and SDNN (standard deviation for calculating the total RR interval), a frequency domain characteristic parameter LF/HF ratio (low-frequency power/high-frequency power ratio) and a nonlinear characteristic parameter sample entropy, wherein the EMG characteristic parameters comprise a root mean square value RMS of the amplitude of an electromyographic signal, total power or frequency spectrum energy in a preset