| 引用本文: | 郭清,夏虹,韩文伟.物联网的压水堆CRDM故障信息融合方法分析[J].哈尔滨工业大学学报,2015,47(3):83.DOI:10.11918/j.issn.0367-6234.2015.03.014 |
| GUO Qing,XIA Hong,HAN Wenwei.Research of PWR CRDM fault information fusion method based on IoT[J].Journal of Harbin Institute of Technology,2015,47(3):83.DOI:10.11918/j.issn.0367-6234.2015.03.014 |
|
| 摘要: |
| 针对核反应堆一回路堆芯控制棒驱动机构可能发生卡棒、漏棒和滑棒故障,提出了以物联网为决策诊断框架的粗糙集神经网络融合算法,应用MEMS传感器、ZigBee模块和Multi-Agent模块构建了物联网CRDM故障识别系统的感知层、网络层和支撑层,选取6种输入特征代表棒位位移范围,连续属性约简后作为神经网络Agent模块前置输入,将粗糙集Agent模块简约规则作为隐藏层判断准则. 实验结果表明, Multi-Agent诊断结果与实际故障相符,从CRDM故障识别角度验证了物联网应用于核动力装置故障诊断的可行性及粗糙集神经网络融合算法的准确性． |
| 关键词: 物联网 压水堆 CRDM 故障识别 |
| DOI：10.11918/j.issn.0367-6234.2015.03.014 |
| 分类号:TP18 |
| 基金项目:国家自然科学基金(51379046). |
|
| Research of PWR CRDM fault information fusion method based on IoT |
|
GUO Qing1,2,XIA Hong2,HAN Wenwei2
|
|
(1. Engineering Training Center, Harbin Engineering University, 150001 Harbin, China; 2. National Defense Key Subject Laboratory for Nuclear Safety and Simulation Technology(Harbin Engineering University),150001 Harbin,China)
|
| Abstract: |
| Aiming at the problems of the faults of NPP’s control rod drive mechanism, a fault information fusion method is proposed by using rough set neural network in making decisions based on the Internet of Things. The perception layer, network layer and supporting layer were constructed by involning the MEMS sensor, ZigBee module and Multi-Agent module, and also 6 kinds of input features were selected to represent scope of control rods displacement, reducing attributes as neural network Agent prior input module, module contracted rules as judgment criterion of hidden layers, concise rules of rough set Agent as judgment criteria of hidden layer, respectively. Simulation results verified the feasibility of IoT CRDM rod position monitoring system and the accuracy of rough set neural network fusion algorithm. |
| Key words: IoT PWR CRDM failure to identifying |
