Simultaneous tasks planning and resources assignment in maintenance scheduling under uncertainties

Effective maintenance scheduling and timely execution of maintenance tasks within the given time duration are important to system safety and reliability. In practical maintenance, the practicality of task planning is essential due to uncertainties arising from the actual maintenance environment, limited maintenance operation space, execution challenges, and equipment constraints. This study focuses on enhancing maintenance planning by addressing uncertain factors, evaluating cost and risk, constructing a decision model, and incorporating risk assessment for interactive decisions. It emphasizes the adaptability of maintenance planning through learning and evolution based on historical planning. Unlike previous approaches assuming stable tasks, this study acknowledges that unforeseen changes may occur, necessitating immediate repairs. Thus, a rolling optimization approach is introduced, allowing priority adjustments and dynamic task planning when maintenance resource uncertainty occurs. As one of the critical constraints, the space conflict among maintenance tasks is considered. Numerical experiments are conducted with 12 certain tasks and 7 potential tasks to show the optimal solutions with different uncertain scenarios, and case studies verify the optimality of the solutions.