Title :
Distributed Resource Management Using Active Supervisory Predicate Control
Author :
Tomlinson, Alex I. ; Hoagland, Greg M. ; Garg, Vijay K.
Author_Institution :
Department of Electrical and Computer Engineering, The University of Texas at Austin, Austin, Texas 78712
Abstract :
We model the problem of Distributed Resource Management using supervisory predicate control, and develop a supervisor to manage the resources according to a desirable predicate. The desirable predicate states that resource requests (uncontrollable event) must be serviced according to some cost constraint. Traditional methods for developing the supervisor fail because an unconstrained sequence of uncontrollable events will always lead to a state in which the desired predicate does not hold, hence the controllable predicate would be false. We place a limit on the number of consecutive uncontrollable events for which cost-effective service is guaranteed, and define a dynamic controllable predicate. The dynamic controllable predicate enables the plant to utilize more of its state space than a static controllable predicate would allow. A dynamic controllable predicate requires the supervisor to actively manipulate the state of the plant using forced events in addition to the usual task of resticting controllable events.
Keywords :
Control systems; Costs; Discrete event systems; Engineering management; Force control; Manipulator dynamics; Research and development management; Resource management; State-space methods; Supervisory control;
Conference_Titel :
American Control Conference, 1992
Conference_Location :
Chicago, IL, USA
Print_ISBN :
0-7803-0210-9