Research On Intelligent Vehicle Driver Model Based On Multi-objective Evaluation Of Steering And Speed Integrated Control

The driver model is the decision-making and control center of intelligent vehicle,usually referred to as "vehicle brain".In order to improve the adaptability of intelligent vehicles under complex driving conditions and the agility of decision-making control system,and to simulate the manipulating behavior of real drivers under the the man-vehicle-road closed-loop system,this thesis proposes the intelligent vehicle driver model based on multi-objective evaluation of steering and speed integrated control.The research was carried out from the aspects of adaptive adjustment of preview time,coupling analysis of longitudinal and horizontal motion control,comprehensive speed decision based on multi-objective evaluation,comparative analysis of simulation experiment and real vehicle test.Firstly,the intelligent vehicle steering control driver model is established based on preview-follower theory,and analyzes the influence of factors such as road environment and vehicle driving state on the preview time of intelligent vehicle driver model.The basic preview time and the compensation preview time are used to reflect the influence of different factors on driver's forward-looking behavior respectively.Combining the basic preview time with the compensation preview time,the adaptive preview time model based on BP neural network is established.Using the Carsim/Simulink co-simulation platform,the adaptive preview time intelligent vehicle driver model is designed,and simulation and analysis are carried out for both normal driving mode and aggressive driving mode.The simulation results show that the proposed adaptive preview time model can effectively improve the path tracking effect of the intelligent vehicle driver model.Secondly,the coupling mechanism between speed model and steering model in driver model is analyzed,and the relationship between the driver model parameters and the vehicle model parameters is established through the error analysis.The parameters of the equivalent vehicle model are identified by the input and output data of the complex vehicle model,and the input signal and the algorithm are optimized according to the identification error results.The relationship between the model parameters and the vehicle speed is fitted by the best fitting method,so that the dynamic characteristics of the intelligent vehicle driver model can be reflected more accurately.Thirdly,in order to simulate the complexity and uncertainty of the driver's speed control behavior and the constraints of the vehicle dynamic characteristics and road conditions on the driver's speed selection,several evaluation indexes are set up to describe the driver's performance characteristics on the vehicle's longitudinal speed control.According to the evaluation indexes,the fuzzy rule base is designed,and the classical fuzzy controller is chosen as the fuzzy logic controller structure for the ideal longitudinal acceleration selection.Then,according to the dynamic characteristics of the vehicle model,the control and correction link of driver model is established.The driver model based on steering and speed integrated control is built in the Carsim/Simulink co-simulation platform and compared with the Carsim normal driving mode.The simulation results show that the driver model based on steering and speed integrated control can effectively simulate the self-tuning behavior of the real driver for vehicle speed control.Finally,a variety of driving conditions are designed and the driving behavior data of skilled drivers under different conditions are collected,then reasonable test data under various driving conditions are filtered,and the coordinate transformation of vehicle location data is carried out.At the same time,in order to make the driving condition in the simulation experiment as consistent as possible with the real vehicle test condition,the vehicle model parameters in the simulation model are modified according to the parameters of the test vehicle.The trajectory of real vehicle test is reappeared on the map the collected data are analyzed and optimized,then the trajectory of the simulation model under each driving condition is established.Through the comparative analysis of the simulation model and the real vehicle test,it is shown that the intelligent vehicle driver model based on multi-objective evaluation of steering and speed integrated control can effectively perform the path tracking process under a variety of road scenarios,and also simulate the steering and speed integrated control behavior of the skilled driver.