Design And Application Of Flexible Tactile Sensors For Driver Fatigue Monitoring System

Drowsy driving is an important cause of traffic accidents.Driver fatigue monitoring systems can issue warnings when drivers are tired,remind them to rest,and thus reduce the probability of traffic accidents,which has become a research focus.However,most of the current methods for monitoring driver fatigue are visual monitoring methods which use camera monitoring devices to distinguish the drivers’ behavioral characteristics(eyes,facial expressions,body movements,etc.).This facial recognition performance is vulnerable to factors such as brightness in the car,ambient light,driver’s eye size,and whether glasses are worn.Visual monitoring methods may reduce performance significantly under interference from sunlight reflection and glasses reflection.Aimed at this problem,many scholars have proposed driver fatigue monitoring systems based on physiological signals.But physiological sensors need to be worn on the driver’s head or attached to the skin,which is invasive and adverse to safe driving.Thus,there are still technical bottlenecks in the high-precision,interference free,and robust driver fatigue monitoring system.With the development of intelligent cockpit technology,human-machine tactile interaction based on tactile information provides new technological ideas for monitoring driver fatigue status.Tactile sensors can directly perceive changes in the driver’s grip strength and sitting pressure with a delay of milliseconds.Compared with visual monitoring devices,they are more direct and sensitive,and have a relatively small impact on the driving experience.In recent years,their application in intelligent car cabins has received widespread attention from all sectors.However,there is currently limited research on monitoring driver fatigue status through tactile perception channels and the sensors used have a single function.However,the cabin tactile environment is complex and diverse,and there are still problems in efficiently preparing flexible tactile sensors with adjustable performance,high flexibility,and rich perception capabilities.This paper,which is relies on the National Natural Science Foundation of China Project "Research on Human-machine Parallel Control Conflict Mechanism and Key Technology of Cooperative Co-piloting for Intelligent Vehicles"(51775235),proposes a design and application of a flexible tactile sensor for driver fatigue monitoring systems.A flexible forcesensing material synthesized from silicone rubber and carbon nanomaterials is developed.Based on the flexible force-sensing material,a flexible tactile sensor with multi-dimensional force-sensing ability is designed and fabricated.Flexible tactile sensor is installed on the steering wheel and seat to sense grip force and sitting pressure.Furthermore,a multi-source information fusion monitoring method is proposed for monitoring driver fatigue.This article aims to achieve precise monitoring of driver fatigue by constructing tactile information perception within the cabin.The specific research content is as follows:Firstly,prepare flexible force sensitive materials based on silicone rubber and carbon nanomaterials and analyze their properties.Analyze the conductive and sensitive mechanisms of force-sensitive materials based on the theory of conductive channels and micro tunneling effects.Selecting silicone rubber and carbon nanomaterials as raw materials,multiple conductive rubber force-sensitive materials with different material ratios were prepared by solution blending method.Characterize the mechanical and electrical properties of the prepared force sensitive materials,and study the effects of the types and filling ratios of carbon nanofillers on the conductivity,elastic modulus,tensile strength,and elongation at break of the force sensitive materials.Observing the microstructure of force sensitive materials through scanning electron microscopy,studying the dispersion of carbon nanofillers and the formation of conductive networks,and verifying the feasibility of the preparation method from a microscopic perspective.Secondly,design and prepare a flexible tactile sensor based on force sensitive materials and analyze the sensor performance.Analyze the working principle of the piezoresistive flexible tactile sensor and design two types of tactile sensor structures,pressure sensing and multi-dimensional force sensing,for the cabin tactile sensing needs.Select electrode and substrate materials for flexible tactile sensors and complete sensor preparation experiments based on layer by layer assembly method.Based on this,a 3D printing preparation process was developed to improve material utilization and preparation efficiency.The sensitivity,stability,hysteresis,and multi-dimensional force perception characteristics of the prepared flexible tactile sensor were tested and analyzed.The results showed that the designed sensor can achieve flexible multi-dimensional force perception,with simple signal processing,good sensitivity and measurement range,and high preparation efficiency.Once again,design and test the driver’s tactile information perception module.To obtain accurate and effective driving tactile data,specific flexible tactile sensor arrangements were carried out in response to the multi-dimensional force perception needs of the steering wheel and the high flexibility,large measurement range,and high sensitivity perception needs of the seats.Adaptive sensor signal conversion and processing modules and multi-channel tactile signal acquisition equipment were selected to complete the construction of the driver tactile information perception module.The driver’s tactile information perception module was tested,and the sensor array was calibrated using a handheld push-pull force meter to collect tactile signals generated by different actions of the driver.The results showed that the designed driver’s tactile information perception module can effectively perceive steering wheel grip force and seat sitting pressure.Then,collect and analyze fatigue driving test data.Establish a fatigue driving test platform based on a driving simulator,and collect test data for fatigue driving tests.Based on the Karolinska sleepiness scale,driver fatigue levels were divided,experimental data was filtered and denoised,and a fatigue driving sample database was established.Simplify multichannel tactile signals into tactile parameters with physical significance,and conduct parameter analysis of driver fatigue characteristics to study the changes in driving behavior caused by driver fatigue.Perform time-domain feature extraction and distribution statistics on tactile and vehicle motion data,and select features significantly related to driver fatigue through independent sample T-test.Finally,establish a multi-source information fusion monitoring algorithm for driver fatigue.Establish a feature-level multi-source information fusion driver fatigue monitoring model based on neural network.Conduct correlation analysis between driver tactile features and vehicle motion features to reduce information redundancy of multi-source features.Based on principal component analysis,the dimensionality of multi-source information features is simplified to improve monitoring efficiency.Multiple feature groups significantly related to driver fatigue are used as inputs for the driver fatigue multi-source information fusion monitoring model and model training is completed.The monitoring accuracy of the multisource information fusion feature input and other feature group inputs is compared,and the results show that the established multi-source information fusion monitoring method for driver fatigue can achieve better accuracy and stability in monitoring driver fatigue.