论文部分内容阅读
首先论述各种状态信息和设备状态之间的对应关系,即模式分类的重要性.然后通过学习,建立故障诊断的神经网络模型,并应用于大型旋转机械的故障诊断.实验研究表明,神经网络能够较好地表达训练样本要求的决策区域,具有较强的分类能力;利用机械振动特征信息进行训练的神经网络对大型旋转机械单个故障有较好的联想能力,其识别效果令人满意,投入现场应用是可行的.
First of all discusses the correspondence between various state information and device status, that is, the importance of pattern classification. Then through the study, the neural network model of fault diagnosis is established and applied to the fault diagnosis of large rotating machinery. Experimental results show that the neural network can well express the decision region required by the training samples and has strong classification ability. The neural network trained by the mechanical vibration characteristic information has better associative ability for single large-scale rotating machinery failure and its recognition Satisfactory results, put into the field application is feasible.