【作者】 罗彬； 曾水平；

【Author】 Luo Bin Zeng Shuiping (North China University of Technology,College of Mechanical Electronical and Engineering,Beijing,100144,China)

【机构】 北方工业大学机电工程学院；

【摘要】 Elman网络是一种两层的反馈型神经网络,他的反馈连接时从它隐层的输出到它的输入端。将Elman神经网络运用到铝电解槽的故障诊断中,选取槽况正常和故障情况下的铝电解槽频谱能量为输入样本,经过反复调整训练确定了隐层神经元的个数。利用MATLAB编写相应程序,最终成功诊断出了相应故障,达到了预期的要求。更多还原

【Abstract】 Elman network is a two-feedback neural network,and his feedback connections from its hidden layer’s output to its input.Elman neural network will be applied to the aluminum cell fault diagnosis,select the normal cell and fault conditions of the aluminum reduction cell energy spectrum as the input sample,and after training,set the number of hidden layer neurons.Use the MATLAB to write the procedures and finally success of the corresponding failure to achieve the initial requirements.更多还原