Beceri ve yetkinliğe dayalı kaynak kısıtlamaları ile çoklu proje çizelgeleme: ısıtma endüstrisindeki ar-ge projeleri için bir vaka çalışması

The resource-constrained multi-project scheduling problem (RCMPSP) considers more than one project and its task executions, precedence relations, availability of shared resources with limited total working capacity. The RCMPSP belongs to the class of hard optimization problems. Companies' intensive research and development activities require multiple projects to be started and performed simultaneously with resources with different skills, which resemble the RCMPSP with shared resources. In this problem, resources with different skills are assigned to the project tasks according to their skill and capability. We have developed discrete-time mixed-integer linear programming, continuous-time event-based mixed-integer linear programming, and constraint programming models for the considered problem. We have developed our models with three objectives: minimizing projects' tardiness, minimizing the total project development cost of projects, and maximizing the total quality score of projects, respectively. We have conducted a computational study of the proposed models with a series of test instances. The results show that the constraint programming model reduces computational time and improves solution quality, with an average of fifty-six percent on sixty test instances relative to the mixed-integer linear programming models due to the reduction in the number of constraints and variables. Finally, constraint programming is applied to a real-life problem, as it shows the best performance in computational study. The real-life problem is solved using goal programming and goal attainment techniques as multi-objective approaches. The changes in objective weights, number of resources, and capability values are analyzed.