Improving Steel Stockyard Planning by Coupling Optimization with Stochastic Simulation

Abstract

Ship building involves complicated production processes in a highly competitive environment. Therefore improving and obtaining more efficient production facilities is getting more and more important. This fact results in the increase of usage of simulation and optimization tools in the industry. However, coupling these two fields of applications is still not common. This paper proposes a new application by coupling two commercial tools present in the market, QUEST by Delmia for Discrete Event Simulation (DES) and MODEFRONTIER for optimization in a case of improving a steel stockyard in a shipyard. A multi-objective optimization is carried out by taking three design variables and aiming to optimize the objective functions. Number of steel plate piles in stockyard, steel plate capacity of each pile and frequency of steel plate arrival to shipyard are selected as design variables. The objectives are minimizing the area used for stocking and minimizing the Work In Progress (WIP). It is suggested that shipyards with different steel processing capacities would require different sizes of stockyards and different frequencies of incoming steel plates. Therefore a layout plan should be made based on running optimization tasks. The findings provide that production cost can be reduced by carrying out proper planning. The fundamental knowledge of coupling optimization and stochastic simulation tools may result in significant reduction of production costs by minimizing storing area and work in progress. The new solution is also valid for other fields of ship building such as block erection or steel processing and may be modified for further applications.