1995 Volume 8 Issue 12 Pages 692-700
In a previous paper, we proposed a design method, based on the optimal regulator theory for type-1 digital servo systems. The method can cope, with parameter uncertainties and time-continued disturbance to guarantee steady-state robustness of servo systems. However, when this method is applied to actual systems, elements of the weighting matrix have to be repeatedly adjusted in a trial and error fashion, until the given performance-index of the transient process is satisfied. In this paper, we propose, a design algorithm for the weighting matrix, which consists of the following two steps.
(1) Express quantitatively the qualitative adjusting-knowledge of the designer by using fuzzy sets and estimate roughly the initial value of the weighting matrix with the fuzzy reasoning method.
(2) Compute precisely the optimal elements in the weighting matrix by using the simplex method.
By means of these steps, the automation of servo system design is accomplished. Finally, velocity control of a DC motor is taken as an example to show the effectiveness of the method proposed in this paper.