为了缓解优质医疗资源分配不均衡问题，提升医疗资源的利用效率，国家正全力推进医联体模式。在医联体模式中，复杂约束下的医疗资源管理和调度面临着更大的挑战，而不依赖人力的智能最优化调度则是解决问题的关键，也是未来的发展方向。本项目将面向医联体模式，研究多医院手术室多种医疗资源的协同调度问题。针对确定性情境，提出了双层优化模型，研究工作包括：1）设计数学启发式算法求解上层手术分配问题；2）设计松弛-修复启发式和迭代局部搜索算法求解下层手术排序问题；3）设计上下层的反馈机制，通过迭代不断优化调度方案。针对不确定性情境，采用前摄式-反应式调度框架，研究工作包括：1）设计仿真优化算法求解上层手术前摄式分配问题；2）设计重排序算法求解下层手术反应式排序问题。本项目在理论上探索医联体模式下医疗资源协同调度技术，实践上为医院管理者统筹调配医联体内关键资源提供决策支持，以实现医院、医生及患者的利益最大化。

China has been fully promoting the hospital alliance model to alleviate the allocation imbalance and improve the utilization of quality medical resources, which brings greater challenges for managing and scheduling medical resources subject to various types of constraints. Intelligent optimization and scheduling without relying on manpower is the key to tackling this challenge and becomes the direction of future development. In this project, we propose to study the collaborative operating room scheduling problem involving multiple hospitals, subject to multiple resource constraints, making the objective consistent with China's hospital alliance policy. For the deterministic environment, we plan to develop a bi-level optimization model. In particular, we will develop a matheuristic algorithm for solving the surgical case assignment subproblem at the upper level and a relax-and-repair heuristic along with an iterated local search for solving the surgery scheduling problem at the lower level. The lower and upper level solution schemes and outputs generated are combined using an effective feedback mechanism to achieve interaction between the subproblems at both levels, yielding an iterative mechanism with superior outcome. For the stochastic environment, we plan to use a proactive-reactive scheduling framework to generate robust schedules. In particular, we will develop a simulation optimization algorithm for solving the proactive surgical case assignment problem at the upper level, and a rescheduling algorithm for solving the surgery reactive scheduling problem at the lower level. The project will deliver integrated optimization modeling and scheduling algorithms for the specific problems identified in the project. This also provides immediate decision support tools for hospital administrators to coordinate the allocation of key resources within the hospital alliance to maximize the interests of all stakeholders including hospitals, doctors, and patients.