基于动态推理链的电网故障诊断方法
发布时间：2018-08-27 10:58
【摘要】：研究了结合贝叶斯网络的动态因果推理链故障诊断模型。已有基于推理链的电网故障诊断方法存在推理子链个数较多、构造较复杂、未考虑保护和断路器缺失、错误等不足,针对每个疑似故障元件根据保护逻辑建立原始的动态推理链,对误报、漏报的报警信息进行纠正,形成改进后的动态推理链。对推理链中各保护和断路器动作的可信度进行评估,将各保护和断路器动作信息从二值逻辑的0或1模糊化为0,1之间的可信度值。针对每个疑似故障元件的推理链,建立相应的贝叶斯网络模型,通过贝叶斯反向推理得到元件的故障概率。典型电网的多个故障案例验证了所提方法的有效性。
[Abstract]:The fault diagnosis model of dynamic causal reasoning chain combined with Bayesian network is studied. The existing fault diagnosis methods based on inference chain have many disadvantages, such as the number of subchains of reasoning is more, the structure is more complicated, the protection and circuit breaker are missing, the fault is wrong and so on. For each suspected fault element, the original dynamic inference chain is established according to the protection logic, and the false positive and false alarm information is corrected to form the improved dynamic inference chain. The reliability of each protection and circuit breaker operation in the chain of reasoning is evaluated, and the operation information of each protection and circuit breaker is blurred from 0 or 1 of binary logic to the reliability value between 0 and 1. According to the inference chain of each suspected fault element, the corresponding Bayesian network model is established, and the fault probability of the component is obtained by Bayesian reverse inference. The effectiveness of the proposed method is verified by a number of fault cases in a typical power grid.
【作者单位】： 西南交通大学电气工程学院;
【基金】：国家自然科学基金项目(51377137)~~
【分类号】：TM732
本文编号：2207067
[Abstract]:The fault diagnosis model of dynamic causal reasoning chain combined with Bayesian network is studied. The existing fault diagnosis methods based on inference chain have many disadvantages, such as the number of subchains of reasoning is more, the structure is more complicated, the protection and circuit breaker are missing, the fault is wrong and so on. For each suspected fault element, the original dynamic inference chain is established according to the protection logic, and the false positive and false alarm information is corrected to form the improved dynamic inference chain. The reliability of each protection and circuit breaker operation in the chain of reasoning is evaluated, and the operation information of each protection and circuit breaker is blurred from 0 or 1 of binary logic to the reliability value between 0 and 1. According to the inference chain of each suspected fault element, the corresponding Bayesian network model is established, and the fault probability of the component is obtained by Bayesian reverse inference. The effectiveness of the proposed method is verified by a number of fault cases in a typical power grid.
【作者单位】： 西南交通大学电气工程学院;
【基金】：国家自然科学基金项目(51377137)~~
【分类号】：TM732
【相似文献】
相关硕士学位论文 前2条
1 张楠;基于推理链的电力系统故障诊断方法研究[D];上海交通大学;2015年
2 韩迎春;基于灰色关联度分析和改进推理链的电网故障诊断[D];西南交通大学;2016年
，本文编号：2207067
本文链接：https://www.wllwen.com/kejilunwen/dianlidianqilunwen/2207067.html
