多信息融合车辆行驶状态估计方法研究

Study on Vehicle Driving State Estimation Based on Multiple Information Fusion

准确估计车辆行驶状态信息,可保证汽车主动安全控制。针对目前车辆行驶过程中的状态估计多采用单一集中卡尔曼滤波算法理论上存在实际应用中容错性差、易出现不稳定问题。为提高稳定性,提出了采用联邦容积卡尔曼滤波的车辆行驶状态估计算法。建立了采用Dugoff轮胎模型的非线性三自由度车辆动力学估算模型,将联邦卡尔曼滤波理论和容积卡尔曼滤波理论结合起来设计了联邦容积卡尔曼滤波算法,通过对低成本传感器的信息融合实现对车辆行驶状态的实时准确估计。Car Sim与Simulink联合仿真结果表明,联邦容积卡尔曼滤波算法能够准确估计车辆行驶状态,为实车行驶状态估计提供了新的理论依据。

Vehicle state information is the prerequisite for vehicle active safety control. The single centralized Kalman filter algorithm for vehicle driving state is poor in fault tolerance and stability theoretically. The federal cuba- ture Kalman filter algorithm which combines the federal Kalman filter and the cubature Kalman filter is proposed to estimate the vehicle driving state. The nonlinear 3 - DOF model and Dugoff tire model are adopted as the vehicle driving state estimation model. The Federal Kalman Filter and the cnbature Kalman filter theory are combined to de- sign the algorithm which estimates the vehicle driving state aecurately through information fusion of the low cost sensor signals. The vehicle simulation results based on the CarSim and Matlab/Simulink show that federal euhature Kalman filter theory can accurately and stably estimate the vehicle driving state.