为精确诊断水轮机尾水管涡带,该文提出一种基于小波包特征熵的神经网络故障诊断新方法。对采集到的尾水管压力脉动信号进行三层小波包分解,提取小波包特征熵,然后构造信号的小波包特征熵向量,并以此向量作为故障样本对三层 BP 神经网络...为精确诊断水轮机尾水管涡带,该文提出一种基于小波包特征熵的神经网络故障诊断新方法。对采集到的尾水管压力脉动信号进行三层小波包分解,提取小波包特征熵,然后构造信号的小波包特征熵向量,并以此向量作为故障样本对三层 BP 神经网络进行训练,实现智能化故障诊断。试验结果表明训练成功的BP网络能够很好地诊断机组尾水管是否发生涡带以及涡带的严重程度,为水轮机故障诊断开辟新的途径。展开更多
The present paper describes experimental investigation on the flow pattern and hydrodynamic effect of underwater gas jets from supersonic and sonic nozzles operated in correct- and imperfect expansion conditions. The ...The present paper describes experimental investigation on the flow pattern and hydrodynamic effect of underwater gas jets from supersonic and sonic nozzles operated in correct- and imperfect expansion conditions. The flow visualizations show that jetting is the flow regime for the submerged gas injection at a high speed in the parameter range under consideration. The obtained results indicate that high-speed gas jets in still water induce large pressure pulsations upstream of the nozzle exit and the presence of shock-cell structure in the over- and under-expanded jets leads to an increase in the intensity of the jet-induced hydrodynamic pressure.展开更多
文摘为精确诊断水轮机尾水管涡带,该文提出一种基于小波包特征熵的神经网络故障诊断新方法。对采集到的尾水管压力脉动信号进行三层小波包分解,提取小波包特征熵,然后构造信号的小波包特征熵向量,并以此向量作为故障样本对三层 BP 神经网络进行训练,实现智能化故障诊断。试验结果表明训练成功的BP网络能够很好地诊断机组尾水管是否发生涡带以及涡带的严重程度,为水轮机故障诊断开辟新的途径。
文摘The present paper describes experimental investigation on the flow pattern and hydrodynamic effect of underwater gas jets from supersonic and sonic nozzles operated in correct- and imperfect expansion conditions. The flow visualizations show that jetting is the flow regime for the submerged gas injection at a high speed in the parameter range under consideration. The obtained results indicate that high-speed gas jets in still water induce large pressure pulsations upstream of the nozzle exit and the presence of shock-cell structure in the over- and under-expanded jets leads to an increase in the intensity of the jet-induced hydrodynamic pressure.
