Study On Nonlinear Observer Method For Vehicle Tire Force And Velocity Estimation

In order to reduce trafc accidents, along with the development of information tech-nology in the automotive feld, the demand of vehicle handling stability and active safetyis increasing. Many domestic and foreign researchers have already developed a variety ofefective automotive electronic control system to adapt to the intellectualized trend. Asthe automotive electronic control system is mostly based on the vehicle state informationto implement appropriate control logic, it has become a major problem to acquire vehicledriving state parameters information accurately and real-timely to achieve the control ofvehicle active safety. For now, the vehicle state information is usually measured directlythrough the on-board sensors. But due to the complexity of vehicle driving condition,the low measurement accuracy and production costs of on-board sensor etc, it is im-practical to simply use on-board sensors to obtain all the state information. Accordingto the vehicle state estimation theory, it is the research focus to use standard onboardsensor information to estimate vehicle driving state parameters to improve vehicle safetytechnology performance.The achievement of vehicle acceleration, deceleration, cruise and steering controlrelies on the driving friction between road and tire and vehicle driving speed. And itis an important basis to acquire vehicle tire forces and velocities information accuratelyand real-timely to achieve vehicle motion control, driving on a specifc path or warningin a dangerous situation. But these states can not be measured directly. Or thoughthey can be measured, the instruments are much more expensive. In response to theseproblems, combined with the research needs of laboratory project, and considering thestrong nonlinear of vehicle structure with the strong coupling of each state, this paperhas designed a simple, real-time and modular vehicle state estimation scheme. And theverifcation is based on the obtained actual information of test sedan HQ430states. Thespecifc implementation scheme is as follows: Firstly, design4longitudinal tire forcesobservers utilizing the unknown input observer theory, based on a wheel dynamics model.And the lateral tire forces are estimated utilizing tire dynamics based on Dugof tire model. Secondly, design a modular nonlinear full-order state observer to estimate vehiclelongitudinal and lateral velocities and yaw rate, utilising the upper tire force estimationas input information, based on a7DOF vehicle model. Finally, verify the estimationresults of the designed vehicle state non-linear observer. Taking into account the security,economy and other factors, this paper utilises a simulation dynamic model of autonomousdriving Hongqi HQ430sedan which is modelled based on dynamic software veDYNA. Thetire force nonlinear observer method is verifed compared to Kalman flter method. Andthe verifcation to the integrated estimator is made based on this simulation model insteadof real vehicle with diferent driving conditions. And the compared estimator simulationresults are given when the vehicle mass or road surface friction coefcient is changed.Before simulation, this paper has given the vehicle’s simulation parameters and theselection progress of designed observer gains. In addition, deal with several groups ofexperiment data of the real vehicle typical working condition as the input information andcompared data, and further verify the actual estimation results of the designed integratedestimator based on national regulations and combined practice. Analysis and comparisonis made on the estimation results. The simulation results show that the estimation methodhas certain validity and accuracy.