CAJ | 학술논문

针对传统情景树只能产生单一优化解的缺点, 提出了一种多目标情景树算法. 运用Pareto进行多目标排序,并从历史复杂情景中找到与问题情景相似的非支配解集, 从而避免了传统单组权值计算产生的单一优化. 相对于传统情景树算法, 该方法得出的非支配解集更接近问题情景, 更满足实际需求. 通过居民健康管理系统的应用实例, 充分验证了该算法的有效性和实用性.
침대전통정경수지능산생단일우화해적결점, 제출료일충다목표정경수산법. 운용Pareto진행다목표배서,병종역사복잡정경중조도여문제정경상사적비지배해집, 종이피면료전통단조권치계산산생적단일우화. 상대우전통정경수산법, 해방법득출적비지배해집경접근문제정경, 경만족실제수구. 통과거민건강관리계통적응용실례, 충분험증료해산법적유효성화실용성.
Anoptimized algorithm of multi-objective scenario tree is put forward. It finds the non-dominated solution set by using Pareto for sorting multiple objectives,which is similar with the historical complex scenes.The problem of the traditional single optimized way is avoided which calculates similarityby a single weighted set,so the non-dominated solution setgained from the proposed algorithmis closer to the actual demand.Finally,the application exampleof the resident health-management system validates the effectiveness and practicability of the algorithm.