Advances in State Estimation and Controller Synthesis for General Decentralized Control of Discrete Event Systems

Several improvements to decentralized control of discrete-event systems are suggested and investigated. A general form of decentralized control architecure is investigated where controllable events are assigned to be either "permissive" or "antipermissive". The first improvement suggested in this paper is an enhanced state estimator for local general architecture controllers that accounts for past control actions when calculating the set of estimated system states. The new state estimator is applied to the original general decentralized control law to formulate a new control law. The new general decentralized control law, when used to control to a plant, generates a controlled language at least as large as those generated by the original general decentralized control law if a safety condition is satisfied. Two algorithms for generating maximal local control policies are also presented that allow an amount of "steering" of the controlled system. Several properties of these locally maximal control policies are discussed.