基于神经网络的干泵故障识别模型设计要点分析  

Analysis of key points in designing a fault identification model for dry pumps based on neural networks

作　　者：秦柏林 QIN Bolin(China Keyi(Nantong)Semiconductor Equipment Co.,Ltd.,Nantong 226300,China)

机构地区：[1]中科仪(南通)半导体设备有限责任公司,江苏南通226300

出　　处：《中国高新科技》2023年第9期141-143,共3页

摘　　要：神经网络的干泵故障识别模型的设计,需要设计者对神经网络的基本结构,以及具体的故障识别系统结构进行充分了解。具体来说,神经网络故障识别作用的发挥,主要需要借助堆叠字编码达到预期的效果。故障分析的实践也有非常明确的规范性流程。结合堆叠自编码神经网络故障识别功能进行网络模型软件架构的研究,是进一步把握设计要点,确保模型能够为干泵的故障识别提供帮助的重要条件。The design of the dry pump fault identification model of the neural network requires the designers to fully understand the basic structure of the neural network and the specific fault recognition system structure.Specifically,the role of neural network fault recognition mainly needs to use stacked word coding to achieve the desired effect.The practice of fault analysis also has very clear normative processes.The study of network model software architecture combined with the stacked self-coding neural network fault identification function is an important condition to further grasp the design points and ensure that the model can help the fault identification ofdry pump.

关 键 词：神经网络 干泵故障识别 模型设计 堆叠自编码 

分 类 号：TB752[一般工业技术—真空技术]