The resource constrained multi-project scheduling problem (RCMPSP) is a general and classic problem, which is usually considered and solved in a deterministic environment. However, in real project management, there are always some unforeseen factors such as one or more new project arrivals that give rise to intermittent changes in the activity duration (or stochastic duration) of the current project in execution by inserting the new project. This study takes two practical factors in terms of stochastic duration of project activities and new project arrivals waiting for insertion into account of the problem space to form a stochastic resource constrained multi-project scheduling problem with new project arrivals (SRCMPSP-NPA). Based on the benchmark of the PSPLIB (Project Scheduling Problem Library), a new data set is built and 20 priority rules (PRs) are applied to solve the problem and their performances are analyzed. In addition, a heuristic hybrid method is designed for solving the problem timely by dividing the entire scheduling process into multi-state scheduling problems solved by the corresponding rules separately. This approach has been verified by experiments and its performance is better than that of a single rule in most situations.
|Journal||Computers and Industrial Engineering|
|Early online date||12 Sep 2019|
|Publication status||Published - 1 Nov 2019|