Supervisory Control of Discrete-Event Systems via IC3
Paper in proceedings, 2014

The IC3 algorithm has proven to be an effective SAT-based safety model checker. It has been generalized to other frameworks such as SMT and applied very successfully to hardware and software model checking. In this paper, we present a novel technique for the supervisory control of discrete-event systems with infinite state space via IC3. We introduce an algorithm for synthesizing maximally permissive controllers using a generalized IC3 to find (if any exists) a weakest inductive invariant predicate which holds in the initial state, is maintained as the system evolves, and implies safety and control properties. To this end, we use a variation of IC3, called Tree-IC3, as a bug finder to solve the supervisory predicate control problem by iteratively reporting all feasible counterexample traces using a tree-like search, while controlling the system to avoid them. The maximally permissiveness is achieved by finding the weakest of such controllers that is invariant under safety and control properties. Experimental results demonstrate the great potential of using IC3 technique for the purpose of the supervisory control problems.

IC3

Supervisory control theory

Incremental inductive verification

Discrete-event systems

Author

Mohammad Reza Shoaei

Chalmers, Signals and Systems, Systems and control, Automation

Laura Kovacs

Chalmers, Computer Science and Engineering (Chalmers), Software Technology (Chalmers)

Bengt Lennartson

Chalmers, Signals and Systems, Systems and control, Automation

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

03029743 (ISSN) 16113349 (eISSN)

Vol. 8855 252-266

Areas of Advance

Production

Subject Categories

Control Engineering

Computer Systems

DOI

10.1007/978-3-319-13338-6_19

ISBN

978-331913337-9

More information

Created

10/7/2017