Parameter Sensitivity and Redundáncy in a Class of Optimum Time-quantized Control Systems †

The paper describes a numerical method of implementing Chang's statistical design theory for discrete-continuous systems. The method makes extensive use of machine computation for the derivation of the optimum parameter values of the discrete controller and has the main advantage that a quantitative measure of parameter sensitivity is given. The concept of parameter redundancy, whereby a class of sub-optimum systems may be synthesized, is introduced. These systems are only slightly inferior in performance compared with the direct Chang optimum system but have a considerably simplified control configuration.