A slack speed-up based discrete artificial bee colony algorithm for resource-constrained operating room scheduling problem

Abstract

In today’s world, with the rapidly increasing demand for health care and the scarcity of health care resources, it is very important to provide more services under limited conditions. This paper focuses on a resource-constrained operating room scheduling (RCORS) problem with the characteristic of multi-stage. First, a mixed integrated programming model is presented to uncover the impact of no-wait and resources on the problem. Second, a slack speed-up based discrete artificial bee colony (SDABC) algorithm is devised to tackle varying configurations effectively. The solution is represented as a one-dimensional list, from which a complete schedule is derived by a decoding heuristic. Additionally, a speed-up evaluation procedure based on a slack-resource model is developed. In SDABC, an initial population with high quality and diversity is generated based on a problem-specific heuristic. Different strategies are employed in the three stages of SDABC to balance the local optimal, population optimal, and global optimal. A variable neighborhood search is introduced to enhance local exploration capabilities. Finally, we conduct a large number of experiments to verify the performance of the proposed algorithm. Numerical analysis proves that SDABC has significant advantages for solving the RCORS problem.