Aeronautical Research Institute, University of Tokyo
雑誌名
東京大學航空研究所報告
雑誌名(英)
Report/Aeronautical Research Institute, University of Tokyo
巻
26
号
11
ページ
215 - 236
発行年
1960-12
抄録（英）
Sampling and quantization of signals are essential to the digital devices such as digital computers and analog-digital converters. This paper develops statistical optimum design methods of the control systems containing these digital devices under random inputs. In the first part, neglecting the quantization effects, the pulse transfer function of an ordinary sampled-data control system, which minimizes the mean square error, is presented as a function of the spectral density of the reference input signal, under several constraints which come from the physical realizability of the system, time constants and dead time of the controlled process, etc. Although quantizing operation is nonlinear, in statistical treatments it can be linealized as a superposition of a white noise on the signal, and the corrections to the optimum design due to quantization are discussed theoretically. In the later part of this paper, a high speed electronic simulator of sampled-data systems, which was designed and constructed for this study, is described. To examine the theory, experiments of the statistical optimum design were performed using this simulator for several kinds of random inputs. These results are compared with those numerically calculated.
Statistical Design of Control Systems Containing Digital Devices
SA2404854.pdf
(1.81MB)
