FMS control design: an information value approach

The work presented in this paper demonstrates a new approach to design of an information-efficient automatic supervisor of manufacturing systems. Such a supervisor ignores some available information without a significant decrease in the expected performance of the system. In addition to efficient supervisor (space and time), this leads to a design that requires less sensory equipment. The design process itself is made simpler and cheaper. The model of an automated system combines the basic concepts of the theory of controlled discrete event systems with some ideas of the information economics. A constructive algorithm of control design uses heuristic techniques of stochastic optimization. The method is applied to design of a flexible manufacturing cell.