OPTIMISATION OF INSPECTION STATIONS BY USING SIMULATED ANNEALING

Simulated Annealing (SA) is a heuristic solution strategy to find an optimal solution in a problem with multiple local minima by dynamically changing the probability of accepting inferior solutions during problem solving. It is an intelligent approach designed to give a good though not necessarily optimal solution, within a reasonable computation time. SA has received much attention since its introduction and has been applied to a wide range of discrete and continuous optimisation problems. In this paper, the SA algorithm was employed to solve the problems of combinatorial optimisation of the design of inspection systems in multi-stage manufacturing systems. The problem could be addressed as follows: Consider a multi-stage manufacturing system, where there are up to three possible inspection stations to be located at any of ten processing stages, stage 0 to stage 9. The raw material is input to stage 0 and the finished product is output from stage 9. The optimal location will minimise the total cost. The objective function of SA will correspond to the total cost of the manufacturing problems. This paper will describe the representation of the solutions in a binary form, the re-configuration mechanism, and the evaluation of the objective functions based on empirically found cooling schedule. Finally the results of optimisation experiments will be presented and compared to those obtained by Branch and Bound technique found earlier in the literature