The infrastructure systems of urban rail transit are critically important for the safety and stability of train operations. Efficient maintenance on systems requires appropriate allocation of possession times and resources within the limited duration and complicated constraints. An integer programming model is proposed for long-term maintenance scheduling on multi-major systems. Maintenance activities are classified into routine tasks and projects due to their execution cycle, possession duration and influence range. A possession duration indicator is introduced as constraint for the resolution of the time and space possession conflicts. The proposed model aims to minimize the total maintenance cost and penalty cost by arranging maintenance activities on each segment in discrete time periods. The clustered maintenance is applied for the reduction on cost. The results from case study verify the practicability and availability of the proposed model, which is able to provide decision support for long-term maintenance scheduling in urban rail transit systems.

CICTP