The consequence of severe accident can be alleviated by proper severe accident management (SAM). Variety of schemes have been developed to seek optimized SAM strategy for many given accident sequences. In this paper, an optimization platform based on downhill simplex algorithm is developed. Mathematically speaking, the scheme enables automatic finding of optimum of a function in terms of maximum or minimum, which does not require derivatives. The optimization platform is applied to find the optimized depressurization strategy of reactor coolant system aiming at maximizing the reactor vessel failure time for CPR1000 nuclear power plant in a station blackout (SBO) accident. The application process is compared against that with a traditional enumeration algorithm. The comparison shows that the new optimization scheme is about 12 times more efficient than the traditional one, making the new scheme a better candidate to the efficient and appropriate decision-making tool for SAM. It may be applied to other severe accident management study in the future.

可以通过适当的严重事故管理（SAM）减轻严重事故的后果。已经开发出多种方案来针对许多给定的事故序列寻求优化的SAM策略。本文开发了一种基于下坡单纯形算法的优化平台。从数学上讲，该方案可以自动求出最大值或最小值的最佳函数，而无需导数。该优化平台被用于寻找反应堆冷却剂系统的最佳降压策略，旨在最大化CPR1000核电站在停电（SBO）事故中的反应堆容器故障时间。将应用程序与传统枚举算法进行比较。比较表明，新的优化方案的效率是传统方案的12倍左右，这使新方案成为SAM高效，适当的决策工具的更好候选者。将来可能会应用于其他严重事故管理研究。