Elitism-Based Genetic Algorithm Hyper-heuristic for Solving Real-Life Surgical Scheduling Problem

Hyper-heuristic was designed to automate the development of computational search methodologies. Although it has effectively handled a variety of optimisation problems, the surgical scheduling problem (SSP) has not yet been solved by hyper-heuristic. Therefore, we aim to investigate the effectiveness of applying hyper-heuristics to solve a real-world SSP at Hospital Canselor Tuanku Muhriz UKM (HCTM), one of Malaysia’s largest public hospitals. We dealt with daily SSP for block scheduling strategy. The daily SSP arranges the sequence of surgical cases in each operating room (OR) for each day, considering the material and human resource availability. In this work, we aim to maximise the number of surgeries and OR utilisation for a given time horizon while solving SSP. In hyper-heuristic, high-level strategy is heuristic selection mechanism where we applied elitism-based genetic algorithm (E-GAHH). Low level heuristics are problem-specific heuristics where we applied simple move operators. Experimental result demonstrate that E-GAHH approach can provide a more practical and effective schedule with more OR utilisation than HCTM’s solution and traditional genetic algorithm hyper-heuristic (GAHH) approach.