基于并行学习鲁棒自适应的行驶车辆特性参数估计方法研究

CONCURRENT LEARNING-BASED ROBUST ADAPTIVE PARAMETER ESTIMATION OF DRIVING VEHICLE CHARACTERISTICS

针对车辆行驶过程中的特性参数估计问题,基于并行学习思想提出一种鲁棒自适应参数估计方法。通过低通滤波技术,设计一组系统状态和响应函数的一阶滤波变量。结合并行学习,构建特性参数估计的回归向量,并基于参数估计误差向量,设计鲁棒自适应参数更新律。以某型车辆为例,对该方法的有效性进行仿真验证。仿真结果表明,在无/有扰动情形下,该方法均能对车辆质量、粘性摩擦系数、轮胎滚动摩擦系数和空气阻力系数实现良好估计,且能在3 s内实现估计参数的有效收敛,与传统RLS方法相比具有收敛速度快、误差小的特点。

A robust adaptive parameter estimation method is proposed based on the idea of concurrent learning for the estimation of characteristic parameters in the process of vehicle driving.A set of first-order filter variables of system state and response function were designed by using low-pass filter technology.Combining with concurrent learning,the regression vector of characteristic parameter estimation was constructed.Based on the error vector of parameter estimation,a robust adaptive parameter updating law was designed.A vehicle was taken as an example to verify the effectiveness of the proposed method.Simulation results show that the proposed method can estimate vehicle mass,viscous friction coefficient,tire rolling friction coefficient and air resistance coefficient well,and achieve effective convergence of the estimated parameters in 3 s.Compared with the traditional RLS method,it has the advantages of fast convergence speed and small error.