A scenario-based robust optimization with a pessimistic approach for nurse rostering problem

Abstract

Nurse rostering problem (NRP) or nurse scheduling problem is a combinatorial optimization problem that involves the assignment of shifts to nurses while managing coverage constraints, expertise categories, labor legislation, contractual agreements, personal preferences, etc. The focus on this problem serves to improve service quality, nurse health and their satisfaction, and reduction of hospital costs. The existence of uncertainties and inaccurate estimates of the workload leads to a non-optimal or an infeasible solution. In this study, due to the importance of human resource management and crisis management in the health care system, a sustainable approach was developed with a robust scenario-based optimization method. Since NRP is a NP-hard problem, it is impossible to solve it in medium and large sizes in reasonable time. In this paper, a well-known metaheuristic algorithm, namely the differential evolution (DE) algorithm was proposed due to its sound structural features for searching in binary space. Then its performance was compared against the genetic algorithm. The results show that the DE algorithm has good performance.