Research On Mass Identification Based On The Error Adaptive And Road Grade Estimation Based On The Curvature Of Road Vertical Profile

Vehicle mass and road grade generate acceleration resistance and grade resistance,which are both important vehicle parameter and environment parameter.With the improvement in active control system,the online information of mass and grade becomes more important.This thesis focuses on heavy duty vehicle which contains longitudinal acceleration.An error adaptive mass identification model and a road grade estimation method based on the curvature of road vertical profile were built.The ultimate judgment of mass identification method is its accuracy.While the published outcomes cannot adapt the variant driving conditions.When there is much error in both measured data and model parameter,the model cannot make it self-adaptive.Traditional vehicle dynamics model cannot express road information,so the on line road grade estimation method have to accept the assumption that the change rate of grade equals to zero.Road grade changes much slowly comparing to other vehicle states,it is reasonable using this assumption to some extent,but it brings inevitable time lag.In order to get high precise and stable road grade estimation,there are still many problem needed to be solved.In order to improve the precision and stability of mass identification;reduce the time lag in road grade estimation,new attempt has been performed in this research.This paper based on dynamic model building,the purpose is to build mass and road grade estimation model which suitable for practice situation.The main parts of this thesis contains as follows:1)An error adaptive mass identification method has been built.It is necessary to analyze the data in off-line situation before building the real-time model.The structure should be chosen based on the data error.The model proposed in this thesis can self-adapt data error and model parameter error,which tradition methods didn’t take into consideration.The Recurrence Least Square algorithm can both identify the vehicle mass and error in tractive force.2)A road grade estimation method based on the curvature of road vertical profile has been built.Curvature was used to express road grade change rate and was estimated in real-line situation successfully.It can reduce time lag obviously.The road grade estimation method was built by Kalman Filter.The state transition relationship in Kalman Filter is based on vehicle motion and car-following model.Selected states could describe the comprehensive influence of ’driver-vehicle-road’environment.The model do not using longitudinal dynamics,its simple form could well guarantee its stability.3)In order to judge the performance of mass and grade estimation method,fully field test was.carried out.Real road and traffic situation could test the model in a more comprehensive way.A data base was built,which contained off-line calculated road grade.Then we calculated the statistic error of mass and road grade.A correlation coefficient based method was used to calculate the time lag of estimated road grade.The simulation results show that,the mean error of mass identification can be constrained within 10%;for road estimation method,the mean instantaneous error was not bigger than 0.3%,the time lag could be constrained within 0.5 second.