Riassunto analitico
This thesis describes the basic concept of vehicle dynamics modelling and relative simulation for an autonomous 1:10 scale race car present in the laboratory.The aim is to allow a vehicle to follow a predefined path autonomously, acting on the steering command and providing the right acceleration command to achive a reference velocity. Vehicle dynamic model are analyzed, taking into account tire modeling and their nonlinearities. A linearized model is presented and compered with the nonlinear one. One of the problem encountered is to measure the tire cornering stiffness coefficient due to the reduced dimension of the wheel. In order to do that, parameter estimation technique is performed to find a cornering stiffness value that approximate the real one from measurement data. A cruise control system is also implemented in order to allow the vehicle to reach a given reference velocity using the Stanley method. A path following control system is implemented using Model Predictive Control technique, combining longitudinal and lateral dynamics model in order to obtein a longitudinal speed tracker keeping small lateral deviation and yaw angle error. All the simulation in this thesis are performed using Matlab/Simulink.
|
