摘要
针对大容积无缝气瓶安全问题,提出了基于模糊神经网络的气瓶在线监测方法,并利用加速度传感器采集无损气瓶与损伤气瓶的振动信号,且将数据通过小波包分解获取故障特征参数,最后采用模糊神经网络进行模式识别验证方法可行性.诊断结果表明该算法能较好地对气瓶是否损伤做出较准确的判断,因此其适用于气瓶在线监测.
Aiming at the safety problem of large volume seamless cylinder,an online monitoring method based on fuzzy neural network is proposed,and the vibration signal of non-destructive cylinder and damaged cylinder is collected by acceleration sensor,and the fault characteristic parameters are obtained by the decomposition of the data by the wavelet packet,and finally the pattern recognition and verification method is feasible by fuzzy neural network.The diagnosis results show that the algorithm can make a better judgment on whether the cylinder is damaged,so it is suitable for on-line monitoring of gas cylinder.
作者
王振岭
杨智荣
陈祖志
林治国
Wang Zhenling;Yang Zhirong;Chen Zuzhi;Lin Zhiguo(China Special Equipment Inspection and Research Institute,Beijing 100029;Wuhan University of Technology,Wuhan 430070)
出处
《中国特种设备安全》
2020年第6期18-21,71,共5页
China Special Equipment Safety
基金
国家重点研发计划支持项目——2017年国家重点研发计划项目(2017YFC0805605)。
关键词
气瓶
振动
小波包分析
模糊神经网络
Gas cylinder
Vibration
Wavelet packet analysis
Fuzzy neural network
