چكيده به لاتين
With the outbreak of the Covid-19 virus, and the need to maintain social distance, travel demand has decreased significantly. It is not possible to assess the decline in travel demand by simply comparing statistics of previous years; this is because travel demand has had a certain trend over the years, which would have continued in the absence of Covid-19. What should be compared with the demand statistics is the travel demand forecast for the year of the Covid-19 outbreak. This study investigates the effect of Covid-19 pandemic on railway demand in Iran and 10 European countries in two categories of domestic and foreign studies. The domestic study section, used quarterly data from 1390 to 1397. Five models: linear regression, multilayer perceptron neural network, seasonal autoregressive integrated moving average (SARIMA), Holt-Winters and a combination of them, which is actually the average of the results of the five previous models, have been fitted. In the foreign study, only the two mentioned time series models have been fitted. This section, used quarterly data from 2012 to 2018. For identifying the superior model the demand of one year before the outbreak of Covid-19 (normal conditions) was predicted and the results were compared using three factors: MAD, RMSE and MAPE. The superior model in the domestic study is a combination model and in the foreign study it varies. Then, by predicting railway travel demand using the best model, the impact of the pandemic on rail demand was determined. According to the results of this research, three factors include the active population of the country, vehicle per capita, and employment in the polynomial linear regression model are considered significant. Also, in the first year of the Covid-19 pandemic, the railway demand has decreased by 73.12% compared to what should have happened. According to the results of foreign study, in the first three seasons of the Covid-19 outbreak, Ireland with a decrease in demand of 79.68% and Estonia with a decrease of 27.96%, respectively, are ranked as the highest and lowest impact of the pandemic.
