|
| 基于证据贝叶斯网络的核电系统可靠性分析方法 |
| The Reliability Analysis Method For Nuclear Power Systems Based on Evidence Bayesian Networks |
| 投稿时间：2024-12-30 修订日期：2025-02-07 |
| DOI： |
| 中文关键词: 核电系统 可靠性分析 证据网络(EN) GO法 动态贝叶斯网络 |
| 英文关键词:nuclear power systems reliability analysis evidence network GO diagram dynamic bayesian networks |
| 基金项目: |
|
| 摘要点击次数: 3 |
| 全文下载次数: 0 |
| 中文摘要: |
| 核电系统作为双重冗余系统,存在认知不确定性、共因失效以及复杂的动态时序失效行为。传统可靠性分析方法缺乏对时间因素和认知不确定性因素的描述,对核电系统中的可靠性分析过于硬性化。针对贝叶斯网络等传统分析方法的局限性,结合证据理论和动态贝叶斯网络,提出了一种核能系统可靠性区间分析方法。首先在核电系统结构功能的基础上构建系统GO图；然后给出存在认知不确定性时GO图向动态贝叶斯网络的转换方法,以及基于信任函数和似然函数求解顶事件发生概率的方法；进一步,针对共因失效行为,引入β因子模型,解决系统失效逻辑动态性和相关性的重叠问题。最后,运用所提出方法对某余热排出系统进行了可靠性分析,结果表明,该方法得到的系统区间可靠性结果比完全信息条件下得到的更符合实际情况,共因失效的存在会持续扩大系统可靠性评估的不确定性。 |
| 英文摘要: |
| As a dual-redundant system, nuclear power systems exhibit cognitive uncertainty, common cause failures (CCF), and complex dynamic failure behaviors over time. Traditional reliability analysis methods lack the ability to describe temporal factors and cognitive uncertainty, resulting in overly rigid reliability assessments for nuclear power systems. To address the limitations of Bayesian networks and other conventional analysis methods, this paper proposes a reliability interval analysis method for nuclear systems by integrating evidence theory and dynamic Bayesian networks. First, a system GO (Graphical Operator) diagram is constructed based on the structural and functional design of the nuclear system. Then, a conversion method from the GO diagram to dynamic Bayesian networks is presented in the presence of cognitive uncertainty, along with a method for calculating the probability of top event occurrence based on trust and likelihood functions. Further, to address common cause failure behaviors, a β model is introduced to resolve the overlap of dynamic system failure logic and correlations. Finally, the proposed method is applied to the reliability analysis of a residual heat removal system. The results demonstrate that the system reliability intervals derived from this method are more realistic than those obtained under full-information conditions, and the presence of common cause failure continuously expands the uncertainty in the system"s reliability assessment. |
| 查看/发表评论 下载PDF阅读器 |
| 关闭 |
|
|
|