Stability Measurement of Metro Passenger Travel Patterns Based on Information Entropy Model

［Objective］ It is aimed to more accurately measure the individual travel pattern stability of urban rail transit passengers, improve travel prediction accuracy, and address bias issues in existing information entropy models when quantifying stability, providing more reliable data support for transportation planning and management. ［Method］ First, a daily travel chain is constructed based on the sequence of stations visited in a single day, serving as the core data structure for analyzing passenger travel patterns. Next, a similarity measurement method for daily travel chains is designed based on LCS (longest common subsequence) algorithm. The similarity between daily travel chains is evaluated by calculating their LCS length, and a similarity matrix is constructed to present the daily travel repetition degree for individuals. To further enhance the accuracy of stability measurement, a correction coefficient α is introduced to improve the traditional information entropy model, eliminating deviations in the entropy values produced by the base model. Finally, real passenger travel cases are used to construct daily travel chains, and the information entropy values before and after correction are calculated. A comparative analysis is conducted to assess the effectiveness of the model improvements and verify its applicability to different passenger groups. ［Result & Conclusion］ The improved information entropy model effectively eliminates deviations in the entropy values of the base model, providing a more accurate quantification output of the passenger travel pattern stability, adequately adaptable to most passengers. This method offers a new approach to accurately measuring urban rail transit passenger travel patterns, helping to improve travel prediction accuracy and reliability while providing a scientific basis for optimizing transportation operations.