An Examination of Selected Forecasting Models for Projecting LOGAIR Utilization Requirements for the 5Q Route

摘要

LOGAIR is an extensive, high speed transportation network utilizing a combination of military and civilian, air and surface transportation systems to reduce inventory levels and to reduce pipeline times of priority cargo between suppliers and user organizations. This study analyzed the historical data of a single branch of the LOGAIR system (the 5Q route) in an attempt to find a more consistent, reliable, and accurate means of forecasting daily, weekly, and annual LOGAIR base-level airlift requirements. An algorithm was first developed to convert the raw data into a usable format. This data was then analyzed using a number of relatively uncomplicated forecasting methods which are common to most user station computer systems. The models analyzed included the simple moving average, weighted moving average, exponential smoothing, and regression analysis methods. The models were then compared using the mean absolute deviation and tracking signal criteria to identify the best forecasting method. An examination was also conducted to detect any seasonal trends. The results indicated that very little correlation exists between the weight and time variables in the LOGAIR system; however, when these two parameters were considered, the simple moving average provided the best forecast. (Author)