基于扩展卡尔曼滤波的汽车行驶状态估计

Vehicle driving state estimation based on extended Kalman filter

在车辆行驶中,某些状态参量的准确获取对线控转向系统有着重要的作用,但状态参量测量成本高或难准确测量.因此,针对汽车线控转向系统,为了以较低成本获取准确的车辆运动状态,先建立一个三自由度的非线性车辆模型,搭建多传感器网络(转角传感器、加速度计)采集车辆行驶状态;应用扩展卡尔曼滤波理论建立信息融合算法,通过易测的车辆状态信息(转向盘转角、纵向加速度、侧向加速度)融合得出所需的难测车辆状态(横摆角速度、纵向车速);最后搭建仿真平台在双移线工况和角正弦工况下进行仿真验证,并且与无迹卡尔曼滤波算法估计结果进行对比.结果表明,该估计算法能够准确的估计出车辆行驶过程中的状态参数.

Accurate acquisition of certain state parameters plays an important role in the steer-by-wire system during vehicle travel,but state parametric measurements are costly or difficult to measure accurately.Therefore,in order to obtain accurate vehicle motion state at a lower cost for automotive steer-by-wire steering system,this paper first establishes a three-degree-of-freedom nonlinear vehicle model,builds a multi-sensor network(angle sensor,accelerometer)to collect vehicle driving state.Then we use the extended Kalman filter theory to establish an information fusion algorithm.Vehicle status information(steering wheel angle,longitudinal acceleration,lateral acceleration)is combined to obtain the required unpredictable vehicle status(yaw rate,longitudinal speed).Finally,the simulation platform is simulated and verified under double-shifting and angular sine conditions,and compared with the unscented Kalman filter algorithm.The results show that the estimation algorithm can accurately estimate the state parameters during the running of the vehicle.