摘要
为了对自动化电器设备的故障状态进行准确的识别,文章引入了(RBF)神经网络的故障诊断方法。并针对其不能学习新状态类型的缺陷,提出了一种改进的算法。并将该算法应用于电器设备的故障诊断,改进的算法除了能够对已知的状态进行准确的识别外,还能够发现并学习未纳入训练样本集的状态类型,从而具备了新状态类型的识别功能。
A new fault diagnosis method for automated electric appliance based on （RBF） artificial neural network theory is proposed in this paper, And also does the paper present a modified calculation method aiming for the defect that the method above can not study new state type and than use it to electric appliance fault diagnosis. The modified calculation method not only identify the aware state but also identify and find a state type not in the table of training sample and which has the function for identifying new state type.
出处
《组合机床与自动化加工技术》
2005年第12期67-69,72,共4页
Modular Machine Tool & Automatic Manufacturing Technique
