Abstract
The utility of the classical optimization and decomposition methods were examined with respect to actuator/control system design. A method of integrating actuator design with control system design to obtain optimal transient response was explored and found to be effective. A test problem consisting of an electromechanical actuator was decomposed and optimized following the strategy proposed by Sobieski [1,2]. A comparison of optimization results from decomposed and undecomposed actuator/control system models was made. The results from the sample problems showed decomposition to be a potentially valuable tool in the optimization of large scale dynamic systems. Methods of discrete optimization techniques to select catalog values for dc motors and gears were examined. In the problems tested, a variation of an exhaustive search was found to be the most reliable method for handling discrete variables since a relatively small number of combinations was involved. A sucessful method for evaluating and optimizing the time response characteristics of dynamic systems was also developed.
