Surgery Scheduling Optimization Based On Flexible Job-shop Scheduling Under Multi-resource Constraints

With the deepening of the medical system reform, the competition of medical marketbecomes increasingly fierce. Hence it is essential to provide an efficient decision makingsupport in maximization of the cost-revenue in operating room management. As the biggestcost and income center, operating room has a critical effect on the whole performance ofhospital. A reasonable surgery schedule can use the human resource and equipments inhospital effectively, shorten the average length of stay and overtime, and then improve thebusiness performance while reducing the operating costs. Under the open surgery strategy, theresources from individual department are integrated to assign to different surgeries, whichgreatly increases the complexity of surgery scheduling problem. Therefore, research onsurgery scheduling problem under multiple constraints is significant both in theory and inpractice.Considering a complete process of surgery including preoperative and postoperativestage, multiple resource constraints involved and the integration of surgical upstream anddownstream resources, surgery scheduling problem is described as an extended multi-resource constrained flexible job-shop scheduling model in this study. An improved antcolony algorithm in which ants crawl between the inner and outer paths alternately isdesigned based on the feature of surgical scheduling problem. On the basis of raw data fromhospital we build a series of benchmark problems, which are solved respectively by theimproved ant colony algorithm and simulation approach on SIMIO. The contrastive resultsshow that the improved ant colony algorithm can not only shorten the makespan, but alsoimprove the balance of resource utilization. Considering the multiple constraints of nurses inpractical work of hospital, this paper defines the surgery scheduling problem under actualnurses’ constraints and proposes the corresponding mathematical model and the modified antcolony algorithm. A test case from a literature of daily nurse scheduling problem is built andthe nurses’ assignment result by ACO and the literature is compared. The case study showsthat the ACO proposed in this paper achieved good results in several performance criteriawhile satisfying nurse’s various constraints, which indicates the practical applicability of ourresearch. Finally we sort the existing criteria of surgery scheduling from different perspectives,and design an ant colony algorithm based on the Pareto sets to solve the multi-objectivesurgery scheduling problem. The experiment results show that the multi-objective ant colony algorithm can optimize the several performance criteria more balanced, in which the mode ofseveral pheromone values in a path is more suitable to solve multi-objective optimizationproblem.