Predicting Suburban Passenger Traffic Flows Using the Fourier Series

Purpose: Predicting passenger traffic volumes is a complicated task because of the uneven passenger flows in time. This paper proposes a solution to this issue using the Fourier series. Methods: The statistics on the number of passengers transported are presented in a time series that has a fixed period. Each period of the time series is assigned an unknown function that matches the values of the time series in the period under study. The expression of approximate values of unknown functions is achieved by a partial sum of trigonometric Fourier series. The Fourier coefficients for each function are determined by the period number. The coefficients of the function for the forecast period are selected in accordance with the most appropriate trend. Results: The study involved selecting that explains the behavior of a time series during the forecast period with high approximation reliability. The deviation coefficient was 0.94 and the total deviation was 3.5%. Practical significance: The forecasting method described can be used to determine the transported passenger volumes, taking into account the seasonal unevenness of passenger traffic, and mobility plans can be created for the future.