Analogue computers for feedback control systems

FEEDBACK CONTROL DESIGN is passing into a third stage of development as the systems become more complicated, less subject to linear treatment, and more susceptible to noise problems. Before 1934, empirical methods were usually satisfactory since regulators required relatively little accuracy. Thereafter, with war requirements for accurate positioning devices, it was necessary to design servomechanisms carefully so that linear theory could be applied. But the present tendency in feedback control systems is to include the effects of saturation resulting from large-signal input, the irregularities of amplifying elements, and to utilize relay systems and nonlinear elements to obtain a better solution to the control problem. To study such systems, analogue computers are useful; this article reviews some of these systems.