Vehicle Longitudinal And Laterl Velocity Estimation Based On Unscented Kalman Filter

As people pay more and more attention to vehicle active safety, Electronic Stability System (ESP) has become the focus technology of vehicle active safety control system. It not only integrates the function of ABS and TCS, but also advances the stability of vehicle under extreme driving situations, so it became a critical control system for vehicle safety and was widely used. The entire ESP control process is based on the measurement and estimation of the vehicle motional state. The vehicle states are the vehicle longitudinal velocity, lateral velocity, yaw rate and sideslip angle that reflect all kinds of vehicle motion and are important to all the electronic control system. Therefore, researches on how to obtain low-cost, accurate and real-time longitudinal and lateral velocity estimated values for the other states and ESP control system of the car have important engineering significance.By analyzing the traits of various estimation algorithms researchers proposed at home and abroad, a new approach of vehicle motional state estimation by using vehicle dynamics model and state estimation method has been widely used. With this technical route and province and major projects technology support program, this paper gives the estimation research of vehicle longitudinal and lateral velocity in close cooperation with FAW Group Corporation R&D Center. Meanwhile, the proposed algorithm is evaluated by simulation verification under various emergency maneuvers and road conditions. The results show that the estimate algorithm developed is not only suitable for practical application, but also has high accuracy and robustness.The work of this paper is as follows:1. The theory of Kalman filterThe theory of Kalman filter have been studied, and the characteristics of the classical Kalman filter and extended Kalman filter have been analyzed. Above this, the application of the unscented Kalman filter theory have been researched and the optimal estimation of state variables of the nonlinear movement system have been described,and at the same time,the specific process have been showed. This laid the theoretical foundation for the next work to the use of unscented Kalman filter for vehicle longitudinal and lateral velocity estimation.2. The road adhesion coefficient estimate based on UKF filterBased on the using of vehicle sensor information, the combination of nonlinear vehicle dynamics model and Dugoff tire model and the unscented Kalman state observer, the road adhesion coefficient can be estimated online. The estimation of road adhesion coefficient based on the unscented Kalman filter in Matlab/simulink has been achieved, estimated values are compared with simulation values of Carsim. The results showed that the algorithm can estimate the road friction coefficient in real time. This laid the theoretical foundation for the next work to the use of unscented Kalman filter for vehicle longitudinal and lateral velocity estimation.3. The vehicle longitudinal and lateral velocity estimation based on UKF filterBased on the nonlinear vehicle dynamics model, Dugoff tire model and the estimate observer of the road friction coefficient, the vehicle velocity estimation observer has been designed based on unscented Kalman filter. By using sensors information, the vehicle longitudinal and lateral velocity are estimated online in Matlab/simulink.4. Simulation verificationOn the basis of the conclusion of the above study, state estimation of vehicle longitudinal velocity, lateral velocity and road adhesion coefficient have been completed using CarSim7.0 under different road situations and driving situations. By comparing the simulation values from state estimation and Carsim, the feasibility of the algorithm is verified; at the same time, the accuracy and real time of the algorithm is tested and verified by comparing the estimate values with the estimated values of EKF filter algorithm. All of the results have been fairly satisfied, so the algorithm in this paper is proved effective.This paper has researched on the road adhesion coefficient and vehicle longitudinal and lateral velocity estimation algorithm based on the unscented Kalman filter theory. The simulation verification of the combination algorithm of the road adhesion coefficient and vehicle longitudinal and lateral velocity estimate have been made. All of the results show that the estimation observer isn't only suitable for practical application, but also has high accuracy and robustness. So it provides us with a new approach in the research of ESP and other advanced chassis electronic control system.