<![CDATA[This study analyzes the peak movement patterns of passengers on the South Sumatra Light Rail Transit (LRT) system during the 2026 Eid Transportation period using a time series approach. Daily operational data over a 36-day observation period (D-17 to D+17), from March 4 to April 8, 2026, were obtained from the South Sumatra Light Rail Transit Management Agency and PT KAI Divre III Palembang. Data processing was conducted through the Knowledge Discovery in Databases (KDD) framework using Python, Pandas, and Matplotlib. The analyzed variables included the realized passenger volume for 2025 and 2026, the predicted passenger volume for 2026, the number of train trips, seating capacity, and daily load factor. The analysis results indicate that the total realized passenger volume in 2026 reached 484,563 passengers, representing a 3.10% decrease compared to the 2025 realization of 500,067 passengers. The peak service demand occurred on D+2 (March 24, 2026), with 27,485 passengers, a load factor of 81.16%, and a peak index of 2.04, confirming that the highest operational pressure occurred during the return-flow phase, particularly from D+1 to D+4. The Post-Eid Transportation phase contributed the largest share of total passengers at 41.50%. Prediction accuracy produced a Mean Absolute Percentage Error (MAPE) value of 12.05%, indicating that the model was capable of capturing the general direction of passenger demand, although improvements are still needed at extreme turning points. The estimated trend for 2027 indicates a passenger range between 470,026 and 508,791 under three projected scenarios. This study confirms the relevance of the time series approach in understanding the dynamics of rail-based public transportation demand during special operational periods.]]>
