自适应风险规避型院前急救网络设计鲁棒优化

Robust Design of Adaptive Risk-averse Pre-hospital Care Network

为提高突发大规模应急事件下医疗急救系统的响应能力,提出了基于两阶段自适应风险规避型鲁棒优化模型的院前急救网络设计方法。针对不确定的急救需求发生量,通过基于场景的方法构建了最坏情况下的鲁棒条件风险值模型来规避需求未满足的极端风险,确保系统的服务能力。根据线性对偶理论提出等价的鲁棒对等模型以简化求解。最后,基于某市具体数值案例验证了模型的有效性和可靠性。通过对比分析和灵敏度分析发现,自适应模型较风险中性模型和一般随机风险规避模型有更高的需求满足率和更低的损失风险成本,优化方案显著提高了不确定环境下的应急响应效率和需求覆盖率。

In order to improve the emergency response capabilities of the medical emergency system under sudden large-scale demand scenarios,a pre-hospital care network design method based on a two-stage adaptive risk-averse robust optimization model was proposed.For uncertain random emergency demands,a worst-case robust conditional value-at-risk model was constructed through a scenario-based method to avoid extreme risks of unmet demands and ensure the service capabilities of the system.An equivalent robust counterpart was proposed according to the linear duality theory to simplify the solution process.Finally,the validity and reliability of the model were verified based on a specific numerical case in a city.Through comparative analysis and sensitivity analysis,it is found that the adaptive model has a higher demand satisfaction rate and a lower loss risk costs than the riskneutral model and the general stochastic risk aversion model.The optimization planning significantly improves emergency response efficiency and demand coverage under uncertain environments.