多算法结合的汽车参数和状态估计方法研究

为实时准确获取汽车参数及状态信息以提高汽车主动安全性能,提出了一种多算法结合的自适应估计算法。该算法将递推最小二乘算法、蚁群优化算法及容积卡尔曼滤波算法进行有效结合,同时将含有不准确模型参数及未知时变噪声的三自由度非线性整车模型作为标称模型。采用递推最小二乘算法实时估计汽车参数,引入蚁群优化算法实时跟踪容积卡尔曼滤波器的过程噪声及量测噪声,根据目标函数对噪声协方差进行寻优,以解决系统的噪声时变问题,从而获取汽车状态的准确估计。基于CarSim/Simulink的仿真实验结果表明,该算法的状态估计精度高,且具备汽车模型参数校正能力,可以满足系统的控制需要。

Study on Vehicle Parameter and State Estimation Methods Based on Multiple Algorithm Combination

To obtain the vehicle parameter and state information quickly and accurately,and to improve vehicle active safety performance,a new adaptive estimation algorithm combining multiple algorithms is proposed.The algorithm combines the recursive least squares (RLS) algorithm,the ant colony optimization (ACO) algorithm and the cubature Kalman filter (CKF) algorithm,and the 3-DOF nonlinear vehicle model with the inaccurate model parameter and unknown time varying noise is taken as the nominal model.Firstly,the vehicle parameter was estimated by the RLS.Then,the ACO was introduced to track the process noise and the measurement noise of the CKF.Objective function was established to optimize the noise covariance to solve the time varying problems of noise,so as to obtain the accurate estimation of vehicle state.The simulation results based on CarSim & Simulink indicate that the estimation algorithm has high state accuracy,and has good capability to revise the model parameter,which meets the control requirements of the system.