Abstract
Sinusoidal perturbation signal has been usually used in optimizing control systems. But the sinusoidal perturbation causes some difficulties or disadvantages, (1) when the process has unknown or variable dynamic characteristics, or (2) when the simultaneous optimization of a multi-variable process is aimed.
Psuedo-random binary perturbation signals can get rid of these disadvantages due to the sinusoidal perturbation.
In this paper, M-sequence signal is used as the pseudo-random binary perturbation signal, and experimental and theoretical study on the optimization of processes with variable dynamic characteristics and on the simultaneous optimization of two-variable processes is described.
