Machine Loading in Flexible Manufacturing system using Artificial Immune algorithm

Abstract

In this project thesis the FMS loading problem is discussed with the objective to minimize the system unbalance and throughput by the use of Artificial Immune system. Manufacturing technology focuses primarily on flexibility and productivity. With the product variety and product life being the characterizing standards it is important that the flexibility of the job shop is maintained as its efficiency is increases. The complexity of a basic Machine loading problem in FMS is very high due to the different flexibility criteria as Part selection, Operation allocation and the various constraints involved. This dissertation proposes a soft computing technique with constraints on tool capacity and workload of the machine. The aim of using this algorithm is to reach an optimal solution and to ease the tedious computations in large problems involving loading which are NP hard problems. Immune algorithm is a very suitable method due to its self learning and memory acquisition abilities. First some sample machine loading problems are collected from the literature and the optimal system unbalance of the machine is calculated using LINGO optimization software. This project improves some issues inherent in existing techniques and proposes an effective Immune algorithm with reduced memory requirements and reduced computational complexity. The proposed Algorithm is tested on 3 problems adopted from literatures and the results reveal substantial improvement in solution quality over the existing basic mathematical approaches.