Robust Optimization Model For Compositions Of A Fleet Of Containers Deployed Under Demand Nondeterminacy

Affected by the financial crisis, the international shipping market has been down and has lasted for nearly five years since2008. Excess capacity is a fundamental problem that blocks the development of current shipping industry. Imbalance between supply and demand and low freight make shipping companies in trouble. The rising rigid cost of fuel and manpower etc make the enterprises’situation worse. One of the main risks of liner shipping operation is the freight demand fluctuation. Although changing ships in different sizes can accommodate the increase or decrease of the freight demand in shipping market, certain operating risk will exist due to the huge investment of changing ships. In the face of the nondeterminacy of shipping market’ freight demand fluctuation, if the decision of changing ship once makes mistakes, then it may cause the plight of excess or insufficient capacity, even can cause devastating attack for small and medium-sized shipping companies. Therefore, meeting the freight demand fluctuation of liner routes, realizing the fine-tuning of liner shipping capacity by deploying containers reasonably is significant to avoid the risk of shipping and lower down the cost of shipping operation.According to the characteristics of freight demand fluctuation, that is to say, strong seasonality, seasonal fluctuations prediction method was used to predict the future freight demand. A deterministic optimization model for compositions of containers deployed (DOMFCD)in a container shipping line is formulated by using quantities of various containers under different terms during observed period as decision variables, according to the diversity of using-costs and idle-costs of owned containers, long-term leased containers and short-term leased containers. By introducing the theory of robust optimization, a robust optimization model was developed based on the DOMFCD, taking into account of the nondeterminacy of freight demand fluctuation. Making a numerical case of a shipping line conducted as an example, the article used the same period average method, the trendratiomethod and the link relative method to forecast the future freight demand respectively on the basis of previous freight demand dates in shipping line, then used the predicted outcomes as the nondeterministic set of the robust optimization model. Finally the article used the Lingo software programming to solve the model. Through the robustness analysis of the optimization results, and the contrast analysis of deterministic models, the article verified that the proposed robust optimization model was effective and stable for asset deployment in a container shipping line to prevent risks from the freight demand fluctuation.