The Car Test Simulation Based On Adaptive Kalman

To meet the increasing requirements for engine, HIL is increasing used in automotive control system.PID is used in the system control but it is limited. Kalman filter is a kind of estimation,especially on the dynamic system.It can be running on the realtime systemto estimate and predict.one of the most important part in the Kalman filter is state space equation.It is based on least mean-square rule and is precisely decided by the experiences and previousknowledge on the system and it is used not only in one-dimensional random process,but also on the multidimensional system. Incorrect knowledge of noise covariance matrices and system model may cause the system degradation even divergence in the filtering performance.But it is difficult to decide the equation of the system in the engineer application,especially when there are lots of noises.Neural network is a power tool that can realize anything system even if the system is nonlinear.In fact it is a black box that only need to know the input of systemand the output of the system which we call system identification.This paper firstly use RBF to train the inputs and outputs of the system to get the state space equation by using MatlabBased on the state space equation before.This paper proposed Kalman filter combining with the RBF.This paper firstly discussed PID algorithm and get the pros and cons of PID,based on the pros and cons of PID,this paper proposed the Kalman filter which is discussed with the matlab graph showing the result of kalman filter is better than PID,but also gives the cons of kalman filter.Based on the cons of Kalman filter,this paper proposes using RBF to compensate for kalman filter. This paper uses the RBF combining with the kalman method to replace the PID algorithm in vehicle simulation in Matlab.At last,this paper give the model of the vehicle discussing using V type to developt the engine system and vehicle system.Discuss with the subsystem of the vehicle and gives the theoretical equation of the vehicle subsystem.Based on the Vehicle system,using PID to adjust the system comparing with using Kalman and RBF,the result shows that kalman and RBF method can realize the setup and is better than PID.