Research On The Control Strategy Of Intelligent Vehicle Considering Driver’s Characteristics

Under the trend and background of the deepening of automobile intelligence,automobile intelligence has become one of the research hotspots in automobile field.The ultimate goal of automobile intellectualization is to achieve driverless driving under all working conditions,in order to reduce traffic accidents,reduce energy consumption and improve traffic efficiency.However,at this stage,it can not achieve all driving conditions,the driver and intelligent driving system will exist at the same time in the future.The whole vehicle control technology is one of the key technologies of intelligent driving.In the research and design of the whole vehicle control strategy of the intelligent vehicle,the factors of the driver in the vehicle are mainly considered.The different preferences of different types of drivers for the horizontal and vertical driving state are taken as one of the design basis of the control strategy.The whole vehicle control strategy of the intelligent vehicle considering the characteristics of the driver is designed to improve the different types of driving people’s acceptance and trust of intelligent system.The research content of this paper is as follows.In the real campus scene of middle and low speed,the experiment of collecting driver’s driving data is carried out.According to the research content of this paper,we recruited different genders and different ages of licensed test personnel to carry out driving test.According to the normal range and distribution of parameters,we filtered and denoised the collected test data.Through comparative analysis,the maximum and minimum value of longitudinal acceleration and the extreme value of yaw rate are adopted to reflect the driver’s characteristics,and through K-means clustering analysis,the characteristic parameter values of different types of drivers are found,which provides the basis for the design of the controller later.The dynamic model of the vehicle is established.The engine model,clutch model,transmission shift model,brake system model and drive system model are analyzed and established,which provides the basis for the later controller design and simulation,especially for the lower controller.The upper controller based on model predictive control algorithm and the lower controller based on longitudinal dynamics inverse model are designed.In the upper controller,the vehicle model is first discretized,the coordinate system of the reference path is transformed,then the prediction model is designed,and finally the constraint conditions and cost functions are designed according to the characteristic parameters of different drivers.The lower controller calculates the required driving or braking torque through the reverse dynamic equation,and then obtains the required throttle opening and brake pedal pressure through the driving braking switching strategy.The simulation test is carried out for the model and controller combined with the driver type characteristic parameters.First of all,it is verified that the same type of driver can effectively control the vehicle driving under the conditions of straight path and sinusoidal path,different vehicle speed,different road adhesion coefficient,different discrete cycle simulation tests.Then,under sinusoidal condition,the stability of the controller designed under different driver parameters is verified.Compared with the real driver driving data,the designed controller can reflect the driver’s characteristics well.