Institute of Transportation, MOTC

　　In recent years, the cooperation between colleges and the behavior of attending inter-school classes significantly strengthen. In order to service teachers and students for attending inter-school classes, inter-school buses for class attendance arise. However, the inter-school bus route/schedule is manually determined by planning personnel with experience, without optimization from a systemic perspective. Consequently, the effectiveness of the resulting route/schedule is unknown. Therefore, in this research, based on the perspective of the cooperated schools, we develop a bus scheduling model with many-to-many OD demand for inter-school class attendance, subject to the real operating objective and constraints. The model is expected to be an effective planning tool for cooperated schools to decide on a good bus route/schedule for inter-school class attendance. The model is formulated as a special multiple commodity network flow problem, which is characterized as NP-hard. To efficiently solve realistically large problems that occur in practice, we develop a heuristic algorithm by adopting a problem decomposition and collapsing strategy, couple with the use of CPLEX software. Finally, to evaluate how well the model and the solution algorithm could be applied in practice, a case study is performed using the bus operation data for inter-school class attendance from an allied university system in Taiwan. The test results are good, showing that the proposed model and solution algorithm could be useful reference for academics and practices.