Research On Vehicle Path Following Control Based On Human-Vehicle Cooperation

With the improvement of the people's living standard and the upgrade of the consumption concept,the automobile has become the standard means of transportation for the people to travel.The total number of cars has been rising continuously,and the problem of traffic safety has been paid more and more attention by people.Automobile intelligence is an effective way to improve automobile safety.However,from the current level of technology,we still have a long way to go for fully autonomous driving,and smart cars still need drivers for a long time.Smart cars are still inseparable from drivers for a long time.Intelligent driving scheme based on human-vehicle cooperative control will be an important direction of smart cars development in the future.The existing cooperative control methods often ignore the driver's resistance reaction to the segmentation or deprivation of control power,and there is a potential danger caused by the driver's resistance.Therefore,Based on the project of National Natural Science Foundation of China(No.51575223),a new type of steering-by-wire system control mechanism and evaluation method based on driver's characteristics is studied,this paper proposes a human-vehicle cooperative control strategy based on game theory,considering the driver's driving intention and various possible feedback strategies of the driver in the synergic control,and taking following the target path as the scene to carry out the research.It can follow the target path under the cooperative control of man and vehicle.The main tasks are as follows:(1)Samples collection.Combine Labview,CarSim software and NI-PXI hardware to build a driving simulator,build a virtual driving environment for the driver to turn and go straight,collect and process the driver's left and right steering and straight driving data,and use it as the training samples and test samples required driving intention recognition model.(2)Set up the driving recognition model.In view of the shortcomings of Hidden Markov Model(HMM)in identifying driving intention,the genetic algorithm(GA)is used to train the HMM parameters to obtain the global optimal model parameters.The HMM is ignored for the similarity of the recognition samples.In the case of misidentification,the support vector machine(SVM)algorithm is introduced to construct the GA-MGHMM-SVM cascade algorithm to identify driving intentions.(3)Set up the path follow-up autonomous steering controller.The three-degree-of-freedom vehicle model is used to construct the model predictive control(Model Predictive Control,MPC)controller.The constraint conditions are set.The MPC controller is used as the autonomous steering controller,and the joint simulation of CarSim and Matlab/Simulink is adopted.The experiment of double-shifting line and snake running condition controlled by MPC is carried out,and the driver driving contrast experiment of driver is carried out with driving simulator,which shows the effectiveness of MPC in the control of path-following.(4)Construct the human-vehicle cooperative control strategy.Combined with the driver's driving intention,the vehicle-man cooperative control strategy between driver and intelligent driving system is constructed when following the path.Taking the dangerous degree of vehicle driving and the satisfaction degree of driver to vehicle control power as the important decision making quantity of cooperative control,the revenue function of driver and cooperative control system is constructed.