| Summary: | 碩士 === 國立臺灣海洋大學 === 運輸科學系 === 100 === Abstract
In recent years, the government has been developing Taoyuan Aviation City in order to strengthen the capability of aviation in Taiwan, and position the function of Taoyuan International Airport in the future. The final target aims at developing it into the hub airport with a balanced development of origin-destination and transshipment in Eastern Asia. Providing the convenient circulation environment for passengers and goods not only assists the development of aviation industry in Taiwan, but also attracts more tourists and business people to come to Taiwan, which prospers its foreign economic relation, trade, and tourism communication. Furthermore, Taiwan’s role of hub airport in Eastern Asia has become more and more important. Under the complicated but still closely related operational types in aviation industry, understanding the demand not only decreases the investment risk, establishes operational policy, but also plays an essential role in practical operation and forecasting airline passenger volume.
In this study, the Time Series, Gray System Theory and Artificial Neural Network were used to construct and improve airline passengers forecasting, and the variation of time lagged degree was also taken into account to analyze the growth trends of the amounts of airline passengers. The results of this study showed that the Gross National Product, Gross Domestic Product, National Income, Labor Force and Labor participation rate affect the amounts of airline passengers. Moreover, the forecast model is proved to be more accurate after adding the variation of time lagged degree, which apparently increases the predictive ability, and the correlation between variables and variables can also be easy to comprehend in return. According to the comparison between the two latest data in the fourth quarter of 2011 and the first quarter of 2012, the passenger capacity in the original data is very close to the results of the forecasting value. The average error is quite low, and the prediction accuracy is up to 94%. This study expects to provide government agencies, airlines and follow-up researchers an reference in operating strategy when doing research or making investment in the future.
|
|---|
