Glocal Control of Manufacturing Factories by Probability Control of AGV Transportation Mechanism: A Dissipativity Approach

This paper presents a glocal control of manufacturing systems which utilizes swarms of vehicles for transporting the materials in the network of multiple subprocesses. With respect to the increasing in system's size, the stochastic properties of transportation mechanism, and the transportation delay, the following question is non-trivial: How to locally balance the stock levels of the buffers in any process while globally stabilizing the overall networks? To this end, we model the overall system by the feedback connection of a deterministic dynamical system and a stochastic memoryless map. The stability of the system is guaranteed by a proposed dissipativity approach. Agent-based simulations show that both the global and local objectives are attained even if the transportation delay varies much longer than the control period.