טכניון מכון טכנולוגי לישראל
הטכניון מכון טכנולוגי לישראל - בית הספר ללימודי מוסמכים  
Ph.D Thesis
Ph.D StudentSzwarcfiter Claudio
SubjectMultimode Chance-Constrained Project Scheduling Problems:
Formulation, Solution and Testing
DepartmentDepartment of Industrial Engineering and Management
Supervisors PROFESSOR EMERITUS Avraham Shtub
ASSOCIATE PROF. Yale Herer


Abstract

Real-life industrial projects are fraught with uncertainties, many times leading to project overruns. In recent years, researchers have been increasingly interested in the development of project management frameworks for coping with uncertainty. This doctoral research investigates three yet to be studied project management and scheduling problems in a stochastic activity duration setting, aiming to find stable baseline schedules. The first problem is the multimode lean project scheduling problem, wherein a project value function must be maximized subject to due date and budget chance constraints. The second problem is the multimode chance-constrained critical chain buffer management problem, in respect of which the project delivery date has to be minimized subject to resource constraints and duration chance constraints. The third problem is how to calculate the tradeoff between the project value and its net present value (NPV) in a resource-constrained multimode setting. One challenge in the third problem is the nonregular objective function: some activities have positive cash flows associated with them, and others, negative ones. Thus, it is desirable to schedule the former as early as possible, and the latter as late as possible. To resolve the three problems, we developed mathematical programming models?which, in some instances, have been solved to optimality?and competitive reinforcement learning based heuristics, checked against established benchmarks.