| 引用本文: | 符杨,苗育植,黄玲玲,刘璐洁,魏书荣,张智伟.基于改进贝叶斯网络的风电机组动态可靠性评估[J].电力自动化设备,2022,42(11): |
| FU Yang,MIAO Yuzhi,HUANG Lingling,LIU Lujie,WEI Shurong,ZHANG Zhiwei.Dynamic reliability evaluation of wind turbine based on improved Bayesian network[J].Electric Power Automation Equipment,2022,42(11): |
|
| 摘要: |
| 传统的贝叶斯网络方法对于新型智能化风电机组获取的“多元异质”状态信息、机组部件故障与状态信息间的耦合关联描述不全面，易造成可靠性评估结果不准确。为此，建立一种融合故障树、云模型及无标度网络的改进贝叶斯网络。引入迭代更新的思路，将改进贝叶斯网络与时序分析方法相结合，提出一种风电机组动态可靠性评估方法。算例结果表明，该方法不仅可以有效地利用风电机组的实时状态信息进行定量的动态可靠度计算，而且可以利用贝叶斯网络、状态结构洞以及无标度网络实现对状态信息与“部件-状态”结构关联特性的表达，提高了动态可靠性评估的准确性。 |
| 关键词: 风电机组 故障树 云模型 无标度网络 贝叶斯网络 状态结构洞 动态可靠性评估 |
| DOI：10.16081/j.epae.202204060 |
| 分类号:TM614 |
| 基金项目:国家自然科学基金资助项目(52177097)；上海市教育委员会科研创新计划项目(2021-01-07-00-07-E00122) |
|
| Dynamic reliability evaluation of wind turbine based on improved Bayesian network |
|
FU Yang1, MIAO Yuzhi1, HUANG Lingling1, LIU Lujie1, WEI Shurong1, ZHANG Zhiwei2
|
|
1.School of Electric Power Engineering, Shanghai University of Electric Power, Shanghai 200090, China;2.Shanghai Donghai Wind Power Co.,Ltd.,Shanghai 200433, China
|
| Abstract: |
| The traditional Bayesian network method is not comprehensive in describing the coupling relationship among the information acquired by the new intelligent wind turbine, such as “diverse and heterogeneity” status information, fault of unit components and status information, which easily leads to inaccurate reliability assessment results. To solve this problem, an improved Bayesian network integrated with fault trees, cloud model and scale-free network is established. By introducing the idea of iterative updating and combining the improved Bayesian network with time sequence analysis method, a dynamic reliability evaluation method of wind turbine is proposed. Case results show that the proposed method can not only effectively use the real-time status information of wind turbine for quantitative calculation of dynamic reliability degree, but also use Bayesian network, status structure hole and scale-free network to realize the expression of correlation characteristics between the status information and “component-status” structure, thus improving the accuracy of dynamic reliability evaluation. |
| Key words: wind turbines fault tree cloud model scale-free network Bayesian network status structure hole dynamic reliability evaluation |
