基于改进粒子群算法的轮胎模型参数两级辨识

为提高PAC89(Pacejka'89 tyre model)轮胎模型的辨识速度和辨识精度,采用加入自适应权重和自然选择性的粒子群算法,并将PAC89轮胎模型参数分为两级,依次进行辨识.以轮胎模型侧偏力曲线的辨识为例,轮胎模型中的刚度因子、形状因子、峰值因子、曲率因子、垂直和水平偏移率为一级参数,通过改进粒子群算法进行一级辨识得到;组成上述因子的特性参数为二级参数,通过改进粒子群算法进行二级辨识得到.一级辨识收敛时的迭代次数小于40,二级辨识收敛时的、迭代次数在100左右,通过实验数据与辨识模型的对比得出平均相对残差为1.6961％.辨识结果表明,采用改进粒子群算法分两级对PAC89轮胎模型进行辨识的方法,能够在保证模型精度的同时提高辨识速度,是一种有效的多参数辨识方法.

Two-stage Parameter Identification of Tire Model Based on Improved Particle Swarm Optimization Algorithm

In order to improve the identification speed and accuracy of Pacejka ''89 Tyre Model , particle swarm optimization algorithm with adaptive weight and natural selectivity is adopted. PAC89 tire model parameters were divided into two levels for identification in turn. Taking the identification of lateral force curve as an example, stiffness factor, shape factor, peak factor, curvature factor, vertical and horizontal offset rate of tire model as first-order parameters were obtained through improved particle swarm optimization. The characteristic parameters of the above factors were secondary parameters, which could be obtained by improving the secondary identification of particle swarm optimization. The number of iterations is less than 40 when the first-order identification converges, and the number of iterations is about 100 when the second-order identification converges. By comparing the experimental data with the identification model, the average relative residual error of identification is 1.6961%. The identification results show that the improved particle swarm optimization is an effective multi-parameter identification method for PAC89 tire model, which can improve the identification speed while ensuring the model accuracy.