A HYBRID EVOLUTIONARY ALGORITHM FOR FMS OPTIMIZATION WITH AGV DISPATCHING

Scheduling is one of the most significant fields in manufacturing system. In this paper, we also consider how to handle materials during searching the optimal solutions of scheduling problem, because its impact in practical flexible manufacturing system (FMS) cannot be ignored. So we used the state-of-art automated guided vehicle (AGV) as a material-handling system in the FMS. We focus on the combination of an operation scheduling, which means to obtain the optimization of manufacturing scheduling, and the routing of AGVs, which is to transport materials of different operations between different machines in FMS. We use network structure to model FMS with AGV system as a material handling system. System constraints and decision variables about FMS especially related to AGVs dispatching can be presented on the network. That is to say that network modeling describes both operation scheduling information and AGV routing path information on a directed network model. We propose a random key-based particle swarm optimization (PSO) algorithm with crossover and mutation operation to avoid premature convergence and to maintain diversity of the swarm. Numerical analyses for case study show the effectiveness of proposed approach comparing with Genetic Algorithm (GA).