Scenario‐oriented repetitive project scheduling optimization

Abstract Research on project scheduling optimization is trending toward many‐objective optimization, in which the number of objectives exceeds three. However, existing studies usually only consider the necessary logical constraints, ignoring descriptions of practical scenarios and corresponding complex constraints, which limit the designed algorithms in problems with such scenarios and constraints. This study focuses on the many‐objective repetitive project scheduling problem considering practical scenarios with complex constraints (MRPSP‐PSCC). Various constraints are described under flexible matching/mapping between multi‐crew, multi‐mode, and multi‐section scenarios. A many‐objective project scheduling model is proposed for synchronous optimization of time, cost, quality, resource usage, and interruption time. The multiphase balance of diversity and convergence nondominated sorting genetic algorithm III (B‐NSGA‐III) with unique advantages for continuous many‐objective optimization problems is transformed for discrete many‐objective optimization. A series of unique designs is employed in the algorithm, including three‐layer coding rules, constraint handling, and local search, to improve the problem‐solving efficiency of the algorithm. The effectiveness and superiority of the model and algorithm for MRPSP‐PSCC were verified through a case study.