Abstract:
Due to the consolidation trends in steel industry and advancements in logistics, steelmakers now face the challenge of choosing the most cost-effective sourcing network for a steel structure project. This research deals with the problem of determining the optimal trade-off between the cost of material supplies and the tardiness cost of a project. The aim of this research study is to provide the mathematical background of the sourcing optimization models for steel structure projects. Another objective is to come up with an efficient solution method for this mathematical model. Due to the stochastic nature of the delivery times of steel components, the optimization problem is solved by stochastic dynamic programming, which is a well-known technique in operations research. The model is applied to an on-going real project carried out in Siberia, and 9.54% and 19.99% savings are obtained in terms of total project cost. A C++ code is developed for the solution of the optimization problem. The required computation time is quite short, which implies that the proposed method can also be utilized on larger project networks. The main research contribution is the formulation of stochastic dynamic programming model that optimizes the cost of sourcing. Furthermore, a fit-for-purpose computer code has been developed for the solution process. In addition, standing on the intersection of civil and industrial engineering disciplines, this study stimulates further studies on project scheduling and cost optimization, and use of linear, dynamic and stochastic programming methodologies on different civil engineering subject matters. Further research can be conducted on the optimization model that is formulated in this study – i.e. resource or capacity constraints can be introduced, optimization can be re-conducted at each stage, partial sourcing for activities can be allowed, continuous probability distribution can be used for activity durations.
