Mark Last

The work builds upon an economic approach to designing an automatic supervisor of a discrete event stochastic process. The supervisory model ignores some available information without a significant decrease in the expected performance of the controlled process. We develop an information-efficient decision system performing as a maintenance supervisor of a combat squadron. A discrete event stochastic model of the controlled system is represented both mathematically and graphically. The problem of information-efficient control of some maintenance activities is formulated and then solved by a heuristic algorithm. We also discuss an efficient use of some computation parameters of the search procedure.