Multiobjective resource-constrained project scheduling with a time-varying number of tasks.

Area of Science:

• Operations Research
• Computer Science
• Artificial Intelligence

Background:

• Resource-Constrained Project Scheduling (RCPS) problems involve tasks with fixed resource limitations.
• Dynamic RCPS problems feature a time-varying number of tasks, adding complexity.
• Previous research introduced McBAR for dynamic RCPS but with limited analysis.

Purpose of the Study:

• To thoroughly investigate the performance of McBAR in dynamic RCPS.
• To evaluate the significance of McBAR's subalgorithms by comparing it with other techniques.
• To extend the scope of dynamic RCPS problem-solving by comparing McBAR with the Estimation Distribution Algorithm (EDA).

Main Methods:

• The study compares McBAR against several other techniques, including the Estimation Distribution Algorithm (EDA).
• Both McBAR and EDA are applied to solve the dynamic resource-constrained project scheduling problem.
• The significance of McBAR's subalgorithms is assessed through comparative analysis.

Main Results:

• The paper extends the investigation of McBAR's solution-searching ability in dynamic RCPS instances.
• McBAR's performance is compared against EDA, highlighting its effectiveness in dynamic scheduling.
• The application of EDA to dynamic RCPS is presented as a novel contribution.

Conclusions:

• McBAR demonstrates significant capabilities in addressing dynamic resource-constrained project scheduling problems.
• The comparative analysis provides deeper insights into McBAR's performance and its components.
• This work advances the field by applying and evaluating EDA for dynamic RCPS, a unique contribution.