Research on Feedforward Parameter Optimization of Linear Servo System Based on Iterative Learning of Orthogonal Projection

Abstract

The method of orthogonal projection is introduced according to the problem of feed forward parameter optimization in linear servo system with high frequency response. At first, orthogonalized basis functions are constructed and the control system is projected onto the axes of the basis functions. Then system parameters are identified along the axes of basis functions by iterative learning and high speed response are compensated by feed forward. The method extends the iterative learning from time domain to the space of orthogonalized basis functions. The simulations and experiments show this method can significantly improve the tracking and high-response performance of linear servo system and meet the high speed and high precision requirements.