|M.Sc Student||Luzon Yossef|
|Subject||Scheduling Appointments via Fluids Control|
|Department||Department of Industrial Engineering and Management||Supervisors||Professor Michal Penn|
|Professor Emeritus Avishai Mandelbaum|
|Full Thesis text|
A deterministic fluid control approach is proposed for modeling a stochastic single server appointment system with time varying mean arrival rates and several types of customers. Optimal control functions of fluid models are identified and proved. The control functions are used to suggest optimal policies which are the basis for a scheduling heuristic. For a deterministic version of the system where constant mean arrival rates are considered, we show that our heuristic is asymptotically optimal for loaded systems. The heuristic is then used for constructing an appointment book for scheduling customers in the original stochastic system. We analyze the performance of the obtained schedule using simulations.