Research On The Dynamic Control Of Intelligent Vehicle Considering Driver Skill

With the improvement of people's living standards,the number of family vehicles is increasing.By the end of 2019,according to the statistics of the transportation administration of the Ministry of public security,the number of civil vehicles in China is 250 million,and the number of motor vehicle drivers has exceeded 383 million.Among them,the driving experience of 9.98% of the drivers are under one year.The main reason for the increase of road traffic accidents is the trend of drivers' transfer to non specialization and the increase of traffic flow.In order to solve these social problems caused by traffic accidents,many scientific research institutions and automobile manufacturers have conducted in-depth research on the system of vehicle dynamic control,developing many advanced chassis control systems and applied them to many mass production vehicles.However,there are significant differences among drivers with different driving skills.They need different perception time,judgment time and action time in the face of complex traffic conditions.At present,most chassis control systems fail to fully consider the influence of driver's skills on vehicle handling and stability.Therefore,in order to solve the problem of vehicle control intervention under the operation of drivers with different skills,this paper studies the characteristics and differences of driving behaviors of drivers with different driving skills,and designs a vehicle dynamics chassis strategy considering the driver's skills.The main contents of this paper are as follows:(1)Driving data collection and analysisIn order to analyze the driving characteristics of drivers,a driver in the loop platform is built based on dSPACE real-time simulation platform and CarSim vehicle dynamics software.A large number of driving data of drivers under two typical testing cycles of double lane change and snake shape are collected,which provides the data basis for the study of driving characteristics of drivers.Through the in-depth analysis of the driver's driving data,the characteristic variables such as vehicle speed,acceleration,lateral position and steering wheel rotation frequency are selected to represent the driver's driving skills.The K-means clustering method is used to realize the unsupervised clustering of driver's driving skills.The driver's driving skills are divided into different categories,which provides the basis for the design of vehicle dynamics control considering different driving skills(2)Driver skill modeling and parameter identificationBased on the collected driver driving data,a steering control model which can represent the driver's skills is established.Using the multi-point preview model proposed by academician Guo Konghui,the driver's operating characteristics are represented by the preview time,neural delay time,arm muscle delay time and the weight coefficient of each preview point in the model.Based on the driver's driving data with different skills,particle swarm optimization algorithm is used to identify the model parameters,and the accuracy of the model is verified by comparing with the driver's driving data.(3)Vehicle dynamics strategy considering driver's skillsA vehicle chassis dynamics control strategy considering driver's driving skills is established.Combined with the driver's steering model and vehicle dynamics model obtained by parameter identification,the decision-making strategy of vehicle additional yaw moment based on model predictive control is established to realize vehicle motion intervention considering driver's driving skills.In addition,the desired longitudinal control force of the vehicle under the driver's control input is obtained by using the sliding mode control strategy.Based on the vehicle dynamics model,a vehicle driving / braking force distribution algorithm is established to achieve the optimal distribution of vehicle control force under the constraints of vehicle actuator.(4)Test verification and analysisFinally,a vehicle dynamics control system considering driver's driving skills is built,and the vehicle control strategy is verified based on the established joint simulation platform of CarSim Matlab / Simulink.The vehicle control results of different drivers under different driving conditions are tested under the conditions of double lane change and snake driving.It is verified that the vehicle dynamics control strategy based on driver's driving skills can effectively improve the vehicle's controllability and adaptability to driver's driving characteristics.