Airplane Passenger Prediction Model at Jendral Ahmad Yani Airport, Semarang

The prediction model for airplane passengers at General Ahmad Yani Airport in Semarang is very important, especially for airport managers, so that the infrastructure at the airport can be facilitated by managers according to the number of passengers available. Airplane passengers in this case include arriving and departing passengers. The prediction model is differentiated between before and after the COVID-19 pandemic, where passenger fluctuations during the Covid pandemic saw a significant decrease in passengers and after normal times there was a significant increase. Due to limited existing data, the independent variable uses GDP, because GDP is considered to represent economic growth and regional economic structure. The airport service area is limited by the presence of other airports close to the airport under study. The model is planned using Artificial Neural Networks (ANN) but if the data cannot be processed then the model is analyzed using Regression. The prediction model will later compare before and after post-Covid. Analysis of existing data shows that the passenger prediction model used in the future is a post-COVID-19 prediction model, the results of which are better applied with other possible independent variables if any.


Background
More accurate passenger prediction models, such as artificial neural modeling can improve estimation accuracy compared to classical modeling [1], and more advanced forecasting approaches such as artificial intelligence techniques to evaluate the performance of neural networks relative to series models time [2], Based on structural time series models (STSM) and time-varying parameter (TVP) regression approaches.This new model, TVP-STSM, is superior to the time series model [3], predicting air transport passengers in airport planning and management is an important stage in determining infrastructure needs [4], the attitudes of intermodal transportation users are influenced by intermodal perceptions [5] and all important variables in the aviation industry will have a potential impact on load factors and the influence of time effects in regression models which require certain formulations [6].Apart from that, it is also necessary to pay attention to the price of airplane tickets, if ticket prices are high then demand for air travel tends to decrease [7].Airports are an important component of the air transportation network, airports have a large impact on the operational performance of the air transportation system, this is due to geographic location, demographics, and economics, airports play different roles in the air transportation system and can spread to other airports, and ultimately can spread the entire network, the performance of this airport has a great impact on the performance of the entire network, [8].So this research was adapted to the existence of other airports in the vicinity to determine the service area

Problem
Problems in the world of aviation in Indonesia, especially Semarang's General Ahmad Yani Airport during the COVID-19 pandemic, which affected the production number of airplane passengers, fell by 63.55%, whereas before the Covid pandemic, the average growth in airplane passengers was 4.36%, and After normal passenger growth reached 25.21%, high passenger growth and limited airport capacity and air transportation passenger services still need to be improved [9].The existence of a new airport around General Ahmad Yani Airport will affect the service area of the airport itself so the service area is also one of the factors that influences the number of airplane passengers.

Service Area
Determining the area served by Ahmad Yani Airport by paying attention to the nearest existing airports and the routes they serve, such as Adi Sumarmo -Solo, Tunggul Wulung -Cilacap and Adisutjipto -Yogyakarta airports [10].Airports are important nodes in the transportation system and can have an important role in supporting the socio-economic development of urban areas, [11].This condition will affect the attraction of the number of airplane passengers at each airport because airplane passengers will tend to choose the closest airport to travel to the same destination.General Ahmad Yani Airport's service area includes several cities/regencies such as Semarang City, Salatiga, Semarang Regency, Kendal, Demak, Kudus, Jepara, Pati, and Grobogan, which can be seen in Picture 1.The green one.

Simple Regression
Simple regression is used because the independent variable is single, the variable in this study uses one independent variable, namely GDP and the fixed variable is the number of passengers, so it is sufficient to use simple regression analysis, where the stages for obtaining the equation in regression include:  The value a is obtained

Number of Airplane Passengers
The total number of airplane passengers at Jendral Ahmad Yani Airport in Semarang from 2007to 2022 can be seen in Table 2. Picture 1. below this.From Table 1. and Picture 1. before the COVID-19 pandemic, the growth in airplane passengers was 4.36%, while during the COVID-19 pandemic, there was a significant decline in the number of airplane passengers from 2019-2020 of 63.55% after the post-Covid pandemic the growth average airplane passengers increased by 25.21%.Later the modeling will be made separately during normal and post-COVID-19 times because it has been tried with 11 direct data, the line graph shows a continuous downward trend, in reality after post-COVID-19 there has been growth in the number of passengers in a positive direction.

Gross Domestic Product ( GDP )
GDP is a type of data to measure and evaluate the extent to which development has been achieved by a region, is a benchmark for regional economic growth and structure, and determines the level of inflation.The GDP data for 2012 -2022 can be seen in Table 3. and Picture 3.  3. and Picture 3. It is known that economic growth in the Ahmad Yani airport service area averaged 4.37%, the decline experienced during the 2020 COVID-19 pandemic was 1.99%, and after the pandemic, there was another growth of 3.98%.

Results and Discussion
Initially, the data will be analyzed using artificial neural network analysis, however, from the process that has been carried out, in 11-year time series data (2012 -2022) by adjusting the amount of existing GDP data, the resulting equation tends to continue to decrease.possible errors in the test samples then tried to sort them using 8 data (2012 -2019) before the Covid pandemic and 3 post-pandemic data, the results were unsatisfactory, and even the existing 3 data (2020 -2022) could not be processed further.For this reason, regression analysis was then tried where in this panel the fixed variable (Y) was the number of airplane passengers and the independent variable was the GDP of the Ahmad Yani Semarang airport service area.The regression analysis process uses IBM.SPSS-26 software.The results of the regression analysis using Anova differentiate the overall data, before and after the COVID-19 pandemic, where Y1,2,3 Number of airplane passengers according to conditions and X1,2,3 GRDP of service areas according to conditions.
The similarity of the analysis results with  Where the linear regression line equation is decreasing, it is not suitable to be used to predict the number of passengers in the future.So this equation cannot be used to predict future airplane passengers, because, in reality, post-pandemic the number of passengers has increased.
The equation of the analysis results with 8 time series data with Y2 as the number of passengers and X2 as GDP, the result is; Y2 = −46256062.475+ 3947077.889.X2, graphic results can be seen in Picture 5.With values R = 0.768, R2 = 0.589, F = 8.615, and t = 2.935, this equation is satisfactory but was not chosen to be used, because post-pandemic growth is quite significant, namely ± 25%, so later it needs to be compared with the third experiment with 3 data.
Then tested using 3 post-pandemic time series data, the results are the following equation: Y3 = −3467419.217+ 12.273 X3, according to the results of Picture 6. below this :

Figure 1 .
Figure 1.General Ahmad Yani Airport Service Area Map.

Table 3 .
Service Area GDP 2012 -2022, 5 Figure 3. GDP of the Ahmad Yani Airport Service Area, From Table