摘  要：具有周期性规律的推力纹波扰动是影响直线伺服系统的高速高精性能的重要原因之一，该文根据推力纹波扰动的周期性规律，在给出直线进给控制器结构及推力纹波数学模型的基础上提出一种基于迭代学习的推力纹波参数辨识及补偿的方法，该方法通过最优性能指标函数确定了迭代学习规律和收敛性，并通过前馈进行推力纹波补偿。实验结果表明该方法显著提高直线伺服系统的位置跟踪精度和高速响应性能，满足其高速高精的要求。
关键词：直线驱动；迭代学习；参数辨识；推力纹波

Abstract: Force ripple is a periodic disturbance which has a negative effect on high speed and high precision of linear drive system, this paper put forward a new method to identify parameters of force ripple disturbance and compensate the force ripple disturbance based on a known control structure and force ripple disturbance model which was established according to its periodic characteristic. This method establishes iterative learning rule and convergent characteristic of parameter identification algorithm according to an optimal performance function and compensates the force ripple disturbance by feedforward. Experimental results show that the iterative tuning method effectively enhances tracking and response performance and meets the high speed and high precision requirements.
Key words: linear drive; iterative learning; parameter identification; force ripple