Showing papers by "W. M. Wonham published in 2002"

Distributed diagnosis for qualitative systems

Abstract: In this paper we propose a novel automaton-based architecture to build a diagnoser, based on which an efficient distributed diagnostic method consisting of local computation and communication is presented. The method proposed here is highly scalable and robust to partial failures of the overall diagnoser.

STCT: An Efficient Algorithm for Supervisory Control Design

Abstract: This paper introduces a new synthesis approach for the supervisory control of discrete-event systems (DES). Our algorithm, named S(mart)TCT after our software package CTCT hitherto in use, is much more efficient than CTCT. Efficiency is achieved by exploiting the modular composition of the plant and specification in DES, and its embodiment in integer decision diagrams (IDDs) as the basic data structure.

Implementation Considerations in Supervisory Control

Abstract: With supervisory control theory it is possible to describe controllers which influence the behaviour of a system by disabling controllable events. But sometimes it is desirable to have a controller which not only disables controllable events but also chooses one among the enabled ones. This event can be interpreted as a command given to the plant. This idea is formalized in the concept of an implementation, which is a special supervisor, enabling at most one controllable event at a time. In this paper, some useful properties are introduced, which ensure, when met, that each implementation of a given DES is nonblocking. The approach is applied to a simple chemical batch process example.

Real-Time Supervisory Control of a Processor for Non-Preemptive Execution of Periodic Tasks

Abstract: In this article, a method for scheduling a processor for non-preemptive execution of periodic tasks is presented. This method is based on the formal framework of supervisory control of timed discrete-event systems. It is shown that, with this method, the problem of determining schedulability and the problem of finding a scheduling algorithm are dual since a solution to the former necessarily implies a solution to the latter and vice versa. Furthermore, the solution to the latter thus obtained is complete in the sense that it contains all “safe” sequences of task execution with the guarantee that no deadline is missed. Examples are described to illustrate this method. Implication of the results and computational complexity associated with this method are discussed.

Robust nonblocking supervisory control of discrete-event systems

Abstract: In this paper, we generalize a notion of robust supervisory control to deal with marked languages. We show how to synthesize a supervisor to control a family of plant models, each with its own specification. The solution we obtain is the most general in that it provides the closest approximation to the supremal controllable sublanguage for each plant/specification pair.

Complexity reduction in discrete event systems

Abstract: The aim of this thesis is to explore new ways to achieve complexity reduction in discrete event systems (DES). We present three different ways of reducing complexity. First, we present a symbolic supervisory control design method for composite systems such that the complete state space never needs to be computed. Instead of a supervisor implemented by a static lookup table, we provide a function that can be efficiently and dynamically computed at each state to determine the control action. This symbolic function can be suitably modified to ensure that the system under control is free of deadlocks. Secondly, we present a heuristic algorithm to reduce the size of a (static) supervisor. Our symbolic supervision scheme is not able to guarantee non-blocking behaviour in the system under control. So to ensure non-blockingness it may be necessary to use a lookup table. Finding the smallest lookup table for a given control task is an NP-hard problem. We propose a greedy supervisor reduction algorithm based on the concept of control covers. This algorithm seems to work quite well in a large number of cases. Finally, we present a compact model of timed discrete event systems (TDES). We use local timers at each state of the TDES to model the passage of time. This model is quite robust to changes in time scale and is closed under control.

Supervisory control of interacting discrete event systems

Abstract: In this paper we present a theory for decentralized supervisory control of a general class of multiprocess discrete event systems referred to as interacting discrete event systems (IDES). An IDES is a discrete event system that consists of a number of components working concurrently and possibly restricted by a well-defined interaction specification. We extend the notion of decentralized supervisory control to this class of systems by adding a supervisor for the interaction specification. Conditions that guarantee the optimality of the decentralized supervision of this class of systems are established and a complexity analysis of the proposed procedure is presented.

Probabilistic reasoning in distributed diagnosis for qualitative systems

Abstract: Su, Wonham et al. (2002) proposed a distributed diagnostic methodology which is highly scalable and robust. Based on the same framework we propose in this paper a probabilistic reasoning method which can be used to improve local diagnoses. We also present a method to determine the optimal next-measurement location.