SARIMA모형을 이용한 김해국제공항의 여객 수요예측

다국어 초록 (Multilingual Abstract)

Air transportation is recognized as a public means of transportation, and the aviation industry is one of the fastest affected tourism industries. However, the aviation industry suffered a lot of damage and faced a crisis due to the COVID-19 outbreak in Wuhan, China in 2019. IATA said that the recovery in aviation demand after COVID-19 will continue throughout 2022, and it is expected to gain further momentum from 2023.
Currently, Gimhae Airport is returning to normal as social distancing and entry restrictions are lifted due to COVID-19. In February 2020, the Gadeokdo New Airport Special Act passed the National Assembly of South Korea, and as Busan became a hot topic as a candidate for the 2030 World Expo, interest in Gimhae Airport increased. Gimhae Airport already has a lot of demand in the Yeongnam region, and demand is expected to increase further when COVID-19 is over, so there is a need to respond to aviation demand.
Air traffic forecasting is essential for airport planning to determine airport facilities and runway systems. In particular, forecasting demand in the aviation industry is used as a reference for economic actors in the industry, research on accurate demand forecasting models is underway. This will help to understand the future forecasting demand and recovery of the aviation industry.
This study attempted to forecast the demand for passengers at Gimhae International Airport and provide policy and practical implications for the aviation industry.
Looking at previous studies in the aviation industry, it can be seen that Box-Jenkins' ARIMA model is used for demand forecast research, which is valuable for the scalability of the ARIMA model. The advantage of the ARIMA model is that the accuracy of the forecast is high, but it is difficult to reflect the periodic characteristics of the time series data. The seasonal ARIMA model, that is, the SARIMA (Seasonal Autoregressive Integrated Moving Average) model, which is a model suitable for time series data with distinct seasonality, was used to supplement this. In this study, the R program, which has the advantage of being able to create and utilize the desired function, was used, and ARIMA(1,1,2)(0,1,0)[12] was obtained to forecast passenger demand. As a result of checking MAPE (Mean Absolute Percentage Error) to verify the accuracy of the model, a value between 0%≤MAPE≺10% appears, and ARIMA(1,1,2)(0,1, 0)[12] model is very reliable. As a result of forecasting passenger demand at Gimhae Airport from November 2022 to December 2030, it can be seen that it shows seasonality and gradually increases over time until December 2030, the forecast period. The demand for passengers at Gimhae International Airport is expected to increase gradually by 2030 starting with a 13.2% increase from the previous year in 2023. Assuming that COVID-19 is over and there are no special events, the demand for passengers at Gimhae International Airport will steadily increase by 2030.
Gimhae International Airport has already developed with the local community and is used by 10 million people a year, and research on demand forecasting at Gimhae Airport is more meaningful at this time to build a new airport in Gadeokdo Island and hold the 2030 World Expo. In the future, it is intended to be basic data for supplementing and expanding facilities at Gimhae Airport, constructing a new airport in Gadeokdo, and planning sales and business of the domestic aviation industry.
Based on this study, it is expected to be used to prepare measures to contribute to the regional development of Busan and the development of the domestic aviation industry, and to provide an opportunity to recognize the necessity and importance of forecasting demand in the aviation industry.

목차 (Table of Contents)

• 제1장 서 론 1
• 제1절 연구배경 및 필요성 1
• 제2절 연구의 목적 3
• 제3절 연구의 구성 4
• 제2장 이론적 배경 5
• 제1절 Forecasting 수요예측 5
• 제2절 관광수요예측 7
• 제3절 항공산업의 여객 10
• 제3장 연구방법 11
• 제1절 김해국제공항 11
• 제2절 분석자료 13
• 제3절 분석방법 16
• 제4장 실증분석 19
• 제1절 정상성 및 계절성 존재 여부 19
• 제2절 모형식별 22
• 제3절 모형추정 22
• 제4절 모형진단 및 검증 23
• 제5절 예측 25
• 제5장 결 론 27
• 제1절 요약 및 시사점 27
• 제2절 연구의 한계 및 향후 연구과제 30
• 참고문헌 32