飞轮电池转子故障诊断系统设计  

Design of Fault Diagnosis System for Batteries Rotor of Flywheel

作　　者：柯友文 汤双清[1,2] 赵晶晶[3] 

机构地区：[1]三峡大学机械与动力学院,湖北宜昌443002 [2]三峡大学新能源微电网湖北省协同创新中心,湖北宜昌443002 [3]三峡大学计算机与信息学院,湖北宜昌443002

出　　处：《三峡大学学报（自然科学版）》2015年第3期81-84,共4页Journal of China Three Gorges University：Natural Sciences

基　　金：国家自然科学基金(51175297)

摘　　要：为了观测飞轮电池转子不对称而产生的失稳、质量不平衡、部件脱落、联轴器连接不平衡等现象,以便及时对故障进行诊断和处理,采用物联网技术设计一个实时对飞轮转子轴心轨迹状态进行评估的诊断系统,利用脉冲耦合神经网络对二维轴心轨迹图进行分割,得到归一化的圆形度时间序列特征向量,在特定的识别器训练和识别,快速地判断工作状态是否正常,使得识别结果更加直观有效.In order to observe the phenomena of rotor misalignment instability, imbalance in quality, parts fall off and coupling imbalance, so that the faults can be diagnosed and processed in time. This paper used inter- net of things technology to design a diagnosis system to evaluate flywheel rotor axis path image state in real time. Using the pulse coupled neural network for two-dimensional axis path image segmentation, a normal- ized circular degree of time series feature vector is obtained which trains and Identifies in particular recognizer, and then quickly determining whether working state is normal or not and making results more effective and in- tuitive.

关 键 词：物联网技术 飞轮转子 脉冲耦合神经网络 故障诊断 

分 类 号：TP206.3[自动化与计算机技术—检测技术与自动化装置]