ABSTRACT
This chapter is devoted to decentralized control schemes for enacting logical constraints such as flow requirements, precedence constraints, non-blockingness, deadlock freeness, no buffer overflows. It discusses a theory of inferencing that allows the design of an as effective control as possible, subsuming the existing frameworks reported in literature. The chapter also discusses a theory that explores the modularity inherent in a distributed system to achieve a computationally efficient design, a type of divide-and-conquer strategy. There are many prior works which deal with decentralized control of discrete event systems (DESs). In decentralized control, multiple local supervisors, each with limited observation and control capabilities, interact to arrive at the global control decisions. The chapter summarizes the results on decentralized control of DESs for enforcing safety and non-blockingness in complex, physically distributed systems such as the manufacturing systems. The existence conditions and the algorithms for control synthesis have been presented for both single global as well as multiple local specifications.
