Exploiting H~~ Sampled-Data Control Theory for High-Precision Electromechanical Servo Control Design

Abstract

Optimal design of digital controllers for industrial electromechanical servo systems using an H <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">infin</sub> -criterion is considered. Present industrial practice is to perform the control design in the continuous time domain and to discretize the controller a posteriori. This procedure involves unnecessary approximations. Alternatively, a direct sampled-data control design can be made, where an optimal discrete time controller is computed for a continuous time standard plant. Though theoretically and numerically solvable, a sampled-data approach is hardly feasible in an industrial environment. In this paper, a discrete time control design approach is taken. Tools that stem from sampled-data control and multirate systems theory are presented to evaluate the intersample behavior in the frequency domain, since intersample behavior is not explicitly addressed during discrete time control design. Experimental results are given to illustrate the effectiveness of the proposed discrete time control design approach and to illustrate the importance of intersample behavior in practical applications