Echelon control for automatic control systems

Abstract

Echelon control has as its goal the control of a single variable through the interaction of a collection of controllers. The objective is that the controllers will work together in a manner such that each has a designated responsibility in overcoming errors of different orders of magnitude or satisfying other system requirements. Echelon control provides a means of relating a number of controllers or actuators so as to use each at its best advantage in a sort of "chain of command" or "echelon" arrangement. In this configuration each echelon is not aware that higher order echelons exist and each echelon sees only lower order echelons and itself in its feedback loop. In this dissertation the general characteristics of an echelon system are set forth. Significant facets are that the stability of the entire system is dependent on the stability of the individual echelons, the type number of the echelon system is the sum of the type numbers of the individual echelons, and the steady state stiffness to constant disturbances is governed by the highest order echelon. Among the varied control specifications that an echelon system can be designed to satisfy individually or in any combination are: 1. High rate of movement of the controlled variable. 2. Ability to follow a step, ramp, parabolic, etc. input exactly. 3. High steady state stiffness to a constant disturbance. 4. System failure protection through redundancy of controllers. This dissertation includes a discussion of these design factors using both theoretically and experimentally obtained results. The experimental system utilized in this research is an angular positional system using two phase servomotors and precision instrument servo components.