Designing appointment system by multi objective simulated annealing

Abstract

In this paper, we consider designing an Appointment System (AS) for an outpatient department of a health care center. Waiting in the line to get consultation from a physician is a major issue in the outpatient clinics, however, minimizing the waiting time would conflict with the number of patients seen by the physician. Therefore, there should be a balance between the objectives of minimizing the waiting time of patients and maximizing the utility of the physician. The AS problem is modeled as a multi objective optimization problem (MOP) with three objectives; minimizing the waiting time per patient, minimizing the number of patients in the clinic and maximizing the doctor utilization. MOP is one of the most challenging field of research because the objectives may conflict of each others. Naturally, there may be no single solution that solves all the objectives. Therefore, there must be trade-off between the objective functions in obtaining a global optimal solution. A two stage procedure is used to construct a Pareto set that contains the non dominated solutions. The simulated annealing (SA) is used in the first stage for exploring the state space. In the second stage, the Pareto set of AS scenarios is constructed based on statistical comparing alternative systems approach with a pre-determined level of significance.