基于改进灰狼算法的电动缸摩擦辨识及补偿

Electric cylinder friction identification and compensation control based on improved gray wolf optimization algorithm

针对电动缸的非线性摩擦难以辨识,且会对六自由度电动Stewart平台(Electrically Driven Stewart Platform, EDSP)位姿的跟踪精度造成不利影响,提出了一种引入Levy飞行和柯西变异的改进灰狼算法(improved Gray Wolf Optimization algorithm introducing Levy flight and Cauchy variation, LCGWO)对电动缸摩擦模型进行辨识,并设计了基于摩擦模型的前馈补偿方法。首先,建立了考虑电动缸非线性摩擦力的EDSP数学模型,并使用Lugre摩擦模型描述电动缸摩擦力。然后,考虑到电动缸的非线性摩擦力难以辨识,提出一种LCGWO同时对摩擦模型中的静态和动态参数进行辨识,并与一般的辨识方法进行对比分析。最后,设计了基于Lugre摩擦模型的前馈补偿算法,并进行实验验证。实验结果表明,相较于其他算法,LCGWO具有更优的收敛速度和预测精度;并且通过基于摩擦模型的摩擦前馈补偿方法对电动缸进行摩擦补偿,可以有效提升平台的位姿跟踪精度。

Nonlinear friction for motorized cylinders is difficult to identify and adversely affects the tracking accuracy of the position of the six-degree-of-freedom Electrically Driven Stewart Platform(EDSP),an improved Gray Wolf Optimization algorithm introducing Levy flight and Cauchy variation(LCGWO)with Levy flight and Cauchy variation is proposed to identify the friction parameters of the electric cylinder,and a feed-forward compensation method based on the friction model is designed.Firstly,a mathematical model of EDSP considering the nonlinear friction of the electric cylinder is developed and the friction is described using the Lugre friction model.Then,considering that the nonlinear friction of the electric cylinder is difficult to identify,an LCGWO is proposed to identify the static and dynamic parameters in the friction model at the same time,and compared and analyzed with the general identification method.Finally,a feed forward compensation algorithm based on the Lugre friction model is designed and experimentally verified.The experimental results show that compared with other algorithms,LCGWO has better convergence speed and prediction accuracy;and the friction compensation of the electric cylinder by the friction feed-forward compensation method based on the friction model can effectively improve the platform′s position tracking accuracy.