Wayan Firdaus Mahmudy
This paper addresses optimization of the integrated part type selection problem and machine loading problem that are considered as NP-hard problems in production planning of flexible manufacturing system (FMS) and strongly determine the system's efficiency and productivity. The integrated problems are modelled and solved simultaneously by using Variable Neighbourhood Search (VNS). A new neighbourhood structure is designed to enable the VNS produces near optimum solutions in a reasonable amount of time. The proposed VNS improves the FMS performance by considering two objectives, maximizing system throughput and maintaining the balance of the system. The resulted objective values are compared to the optimum values produced by branch-and-bound method. The numerical experiments show the effectiveness of the proposed VNS for several test-bed problems.
Faculty of Computer Science, Universitas Brawijaya (UB), Indonesia
