정기선사의 컨테이너 재고 수요예측모델 구축에 대한 연구

This study attempts to establish a precise forecast model for the container inventory demand of shipping companies through forecasts based on equipment type/size, ports, and weekly system dynamics. The forecast subjects were Shanghai and Yantian Ports. Only dry containers (20, 40) and high cubes (40) were used as the subject container inventory in this study due to their large demand and valid data computation. The simulation period was from 2011 to 2017 and weekly data were used, applying the actual data frequency among shipping companies. The results of the model accuracy test obtained through an application of Mean Absolute Percentage Error (MAPE) verified that the forecast model for dry 40’ demand, dry 40’ high cube demand, dry 20’ supply, dry 40’ supply, and dry 40’ high cube supply in Shanghai Port provided an accurate prediction, with 0%≤MAPE≤10%. The forecast model for supply and demand in Shanghai Port was otherwise verified to have relatively high prediction power, with 10%≤MAPE≤20%. The forecast model for dry 40’ high cube demand and dry 20’ supply in Yantian Port was accurate, with 0%≤MAPE≤10%. The forecast model for supply and demand in Yantian Port was generally verified to have relatively high prediction power, with 10%≤MAPE≤20%. The forecast model in this study also had relatively high accuracy when compared with the actueal data managed in shipping companies.