Ship maintenance projects have such characteristics as complex implicated tasks, space interference and uncertain task durations. A mathematical model and optimization algorithm are proposed to solve the stochastic duration optimization problem of ship maintenance.

According to the scenario concept, this paper designs the expected duration as an objective function and constructs a mathematical model, then proposes an improved firefly algorithm to solve the problem. Finally, a group of benchmark projects and one dock maintenance engineering project are carried out to test the validity of the proposed method.

The results show that the proposed method has the best performance in solving the problem. The optimized dock maintenance engineering project has 89.6 days of the expected duration and a 95.6% confidence level. Compared with the original method, the expected duration is reduced by 13.4 days and 13.1%.

This method can provide a basis for planning the schedules of ship maintenance projects.