四轮轮毂电机电动汽车行驶状态估计

Driving State Estimation of Electric Vehicle with Four-wheel-hub-motors

针对四轮轮毂电机电动汽车行驶过程中的状态估计和在数据测量过程中由于偶然因素使观测序列中存在野值的问题,本文中提出了一种基于抗野值鲁棒容积卡尔曼滤波的车辆行驶状态估计算法。首先利用四轮轮毂电机电动汽车的每个车轮的电机驱动力矩容易测得的优势计算轮胎的纵向力,采用Dugoff轮胎模型计算轮胎的侧向力,建立了汽车非线性3自由度车辆模型。接着通过对简单易测低成本传感器信号的信息融合实现电动汽车在行驶过程中的纵向速度、侧向速度和质心侧偏角的准确估计。最后应用Car Sim和Matlab/Simulink联合仿真对估计算法进行验证。结果表明,基于抗野值鲁棒容积卡尔曼滤波的估计算法比扩展卡尔曼滤波估计算法更能较准确地对车辆行驶状态进行估计,且具有较好的实时性。

In view of the problem that in the process of state estimation and data measurement during the driving of four-wheel hub-motor electric vehicle,there exist outliers in observation data sequence due to occasional factors,an algorithm is proposed to estimate the driving states of vehicle based on outliers rejecting robust cubature Kalman filtering. Firstly by taking advantage of the handy measurement of motor driving torque in each wheel of four-wheel-hub motor electric vehicle,tire longitudinal force is calculated,while tire lateral force is calculated by using Dugoff tire model,and a 3-DOF nonlinear vehicle model is established. Then through the information fusion of easily and low-cost measured sensor signals,the longitudinal velocity,lateral velocity and mass-center sideslip angle of the electric vehicle during driving are accurately estimated. Finally the estimation algorithm is verified by co-simulation with CarSim and Matlab/Simulink. The results show that the estimation algorithm based on outliers rejecting robust cubature Kalman filtering can more accurately estimate the driving states of vehicle with better real-time performance,compared with extended kalman filtering.