Date of Award

Spring 1966

Document Type

Dissertation

Degree Name

Doctor of Engineering Science in Electrical Engineering

Department

Electrical Engineering

First Advisor

Joseph J. Padalino

Second Advisor

Frederick A. Russell

Third Advisor

P. A. Fox

Fourth Advisor

Paul H. Goodman

Abstract

A method is presented for the exact determination of absolute and relative stability of linear feedback control systems containing transport or distributed lag. All results are in terms of two variable system parameters. The method utilizes an extension of modern parameter plane techniques that allows for the inclusion of transcendental functions in the system characteristic equation. The design of controllers in linear systems containing transport lag is then considered. A design technique is proposed that allows for the systematic determination of two variable controller parameters in order to meet frequency or time domain design specifications.

The design technique is formulated in terms of the familiar "dominant root" concept for systems that do not contain transport lag. The proposed design technique gives the system designer "at least" as much control over the system response as conventional design procedures for systems without transport lag.

The investigation of absolute and relative stability, as well as the proposed method for controller design, is no more complicated for multiloop feedback control systems than for single loop systems. This is because the characteristic equation of the closed-loop system transfer function is utilized rather than the conventional open-loop methods. Further, if a digital computer is used, high-order systems are dealt with as easily as low-order systems.

A method of constructing the root-locus of systems containing transport lag is then proposed so that this familiar engineering tool can be utilized in conjunction with the proposed analysis and design technique.

Finally, nonlinear systems containing transport lag are considered where describing function analysis is applicable. It is shown that the amplitude and frequency of limit cycles can he predicted where the describing function is real and is dependent upon the amplitude of the input signal to the nonlinearity.

Share

COinS