Model Identification and Synthesis of Discrete-Event Systems

In the discrete-event system (DES) framework, input data are usually given in terms of behavioral descriptions (e.g. transition system and language) and the set of behavioral sequences may be fixed or may be increased in the course of identification by performing new experiments. While only a partial description of the system is assumed in identification, the synthesis problem starts from a complete description of the system. This chapter recalls some basic definitions and notations used in the sequel. In the sequel, it first discusses in detail an approach based on linear integer programming that provides a separating representation of the solution to the synthesis problem in its most general formulation, that is, without any simplifying assumption on the structure and labeling function of the synthesized net. In the second part, the chapter surveys several other similar approaches.