【摘要】 针对抽油机井故障特点,构造了三层小波神经网络,采用带有动量项且学习速率自适应调整的BP算法训练小波神经网络;利用与学习样本和激励函数相联系的方法设置了网络参数的初值,研究了基于小波神经网络的抽油机井故障诊断方法.应用此方法,对某厂的30口故障抽油机井的进行诊断,正确率在95%以上.更多还原

【Abstract】 A method of the fault diagnosis of pumping wells based on wavelet neural networks has been developed. According to the characteristics of pumping wells, a three-layer wavelet adopting neural network has been established. The wavelet neural network was trained with BP algorithm the momentum fraction and self-tuning of learning rate. The initial value of the network concerning the samples and wavelet functions was set. In the practice, the system is used diagnosing a total of 30 faulty pumping wells, the rate of accurate diagnosis is above 95%. The results show that the proposed scheme is very effective.更多还原