| As the development of vehicle electronic, there are more and more vehicle electronic equipments, and their performance are better and better. Among these equipments, a great part of them are used to measure the information of vehicle state and vehicle stability control. The information of vehicle state is the vehicle velocity, wheel spead, yaw, tire forces and tire/road friction that reflect all aspect of vehicle motion. And these years, advanced control devices and assistant systems, which have been suggested aiming to improve the vehicle drivablity and safty are used more and more commonly, such as anti-locking break system(ABS), traction control system(TCS), and electronic stability program(ESP). So the knowledge of the vehicle state is the essential to these systems, and is also the premises of close-loop control. But now, some of the vehicle state information cann't be measured by sensors, because there is no such sensors which can measure them or the sensors are in research or the price of which are so high that they are impossible to be used commonly now. So how to achieve the unknown vehicle state information by the sensors used now is the hot point in research now.Based on 4-DOF vehicle model, we developed a method to estimate tire/road friction using nonlinear observer and exict sensors such as wheel speed sensors, acceleration sensors, yaw rate sensor and hand wheel sensor, and we get a good result.Our work is described as following respects:First, we designed a 9-DOF vehicle model.The 9-DOF vehicle model is made of two parts, tire models and body model. The tire models are the most important parts of the vehicle model, because it has strong nonlinear characteristic, and the nonlinear characteristic of the 9-DOF vehicle model is mostly because of tire models. We looked through a lot of tire models, and finally we chose the Uni-Tire model of Prof. Guo konghui. In body part of the vehicle model, we have considerd the displacement along axis x and axis y, the yaw from axis z, the list from axis x, the pitching from axis y, and the four wheel rotation from their own axis. Finally we combined the tire models and the body model, and get the 9-DOF vehicle model.Sendly, we carried out the tire/road friction estimation method.Based on the 9-DOF vehicle model, we developed the method of the tire/road friction estimation. The tire/road friction estimation method is made of two parts, the saturation tire/road friction estimation and the real-time tire/road friction estimation. Before the estimation, we considered the vehicle model a simple form, and we got a 4-DOF vehicle model. Using feedback, we carried out the method of tire/road friction estimation. At last, we validated the saturation tire/road friction estimation method and the real-time tire/road friction estimation method on the vehicle MATLAB model we have made on Chapter two, and the simulation results showed that our estimation method has a well veracity and real-time ability.Thirdly, we analyzed the stability of the nonlinear observer.In this part, we have analyzed the stability of the nonlinear observer which is used to estimate tire/road friction. We started from a two-dimension nonlinear observer and we have made some assumptions, we concluded that the nonlinear observer is UGAS at origin; Followed the same steps, we have analyzed the four-dimension nonlinear observer, and we rewrited the assumptions. Finally we concluded that based on the assumptions and we choose some proper gains, the four-dimension nonlinear observer is UGAS at origin.As we talked above, we have deeply studied on modeling the 9-DOF vehicle model and tire/road friction estimation method, and we fanalilly got a good simulation result. Our work is in front of using the estimation method on real cars. But there is also some limitation in our method, and we hope in the following works, there will be some improvements in future. |
PDF Full Text Request