Application Research for Train Intelligent Rescheduling Based on the Improved TLBO Algorithm

Abstract

Train rescheduling is a nonlinear optimization problem with multiple constraints. In the view of the theoretical methods and the actual application, the train intelligent rescheduling model is set up according to its characteristics. And the basic teaching-learning-based optimization algorithm has been improved by four optimized mechanisms, such as setting multiple teachers, the self-adaption learning steps, the self-adaption teaching factor and the learning weight. The improved TLBO algorithm with better performance in execution efficiency and solution precision is adopted to solve the train intelligent rescheduling model. Finally, the effectiveness and reliability of the improved TLBO algorithm and the train intelligent rescheduling model has been validated by three simulation examples.