Research On Estimating Algorithm Of Key State Parameters For ESP

Currently, a few key parameters of vehicle state which are very important for Electronic Stability Program(ESP) are cannot be measured by sensors accurately and directly, under the factors of testing techniques and cost of measurement and so on. Therefore, some famous specialists and researchers from the world put forward a series of methods which are estimating the unmeasured vehicle state variables indirectly by the measured one using vehicle dynamic model and kinematic model. In this thesis, the nonlinear vehicle dynamic model with multiple degrees of freedom and the method of nonlinear state estimation are discussed based on the research situation of vehicle state estimation at home and abroad, uniting the special projects of "Developing the vehicle Electronic Stability Program(ESP)" from China Aviation Industry Group. In researching, the vehicle state variables of side-slip angle, longitudinal speed and lateral speed which are significant for ESP are estimated based on UKF algorithm, which can use the nonlinear dynamic model of vehicle with seven degrees of freedom directly, has no need to linearize the model and calculate Jacobeans matrix.In this dissertation, there are some main researching works. Firstly, the estimating model of vehicle side-slip angle is built based on the linear vehicle dynamic model with two degrees of freedom and Kalman filter algorithm, the results of estimation are validated by simulation test and the deviation of estimating results are analyzed. Secondly, the vehicle dynamic model with seven degrees of freedom including longitudinal movement, lateral movement, yaw motion and rotary movement of four tires is built, which provide foundation of vehicle dynamic model for vehicle state estimation. Then, the model of vehicle state parameters estimation is set by using vehicle information such as yaw rate, longitudinal acceleration, lateral acceleration and wheel speed which are easily measured and employing the dynamic model with multiple degrees of freedom built and the estimating algorithm of UKF (Unscented Kalman Filter), which has been applied to estimated the key variables triumphantly and accurately. Finally, the estimating algorithm of UKF is verified by simulation test of software and hardware in the loop truly.