Real Time Control of Stochastic Manufacturing Systems with Jump Semi-Markov

T. Diep, T.-M. Dao (Canada), and S. Abou (USA)


Stochastic control, semiMarkov process, realtime


Manufacturing systems (MSs) are often subject to discrete events in a real-life situation. The control of MSs is a problem due to its complexity and conflicting goals within the different production objective. This paper proposes an optimal control in deterministic horizon for manufacturing systems. Semi-Markov models are used to describe discrete events. The dynamic programming method is implemented and the necessary and sufficient conditions are established as Hamilton-Jacobi-Bellman equation. We make use of the case where simultaneously, the production is kept close to demand at each instant time t while the production cost is minimized. Several analytical and numerical results on optimal policy are presented as hedging point policy.

Important Links:

Go Back