A structured adaptive supervisory control methodology for modeling the control of a discrete event manufacturing system

Two basic measures, model complexity and model construction efficiency, are usually used to evaluate the implementability (or ease of use in practice) of a methodology for modeling the control of a discrete event manufacturing system (DEMS) on the shop floor. Many well-recognized methods are used to represent and analyze the dynamics of DEMS's, but not many relevant applications have been found in developing control software for the shop floor due to their shortcomings in satisfying these two measures. This paper explores a methodology for modeling the control of a DEMS, which leads to ease of control software development, rather than a new representational/analytical tool, by significantly reducing the model complexity (in terms of the number of required control states) and improving the model construction efficiency. First, an extended finite machine, called a deterministic finite capacity machine (DFCM) with parallel computing capability is developed. Based on DFCM's, the complexity growth function of a DEMS control model is linear in the number of synthesized control components. Then, an automaton structure of a DFCM control model, called structured adaptive supervisory control (SASC), is developed. By referring to supervisory control theory, an SASC model is created with three function layers: acceptance, adaptive supervision, and execution. The well-defined structure ensures that the control model can be constructed systematically.