An Intelligent Device Fault Diagnosis Method in Industrial Internet of Things
Dejun Ning, Jingyang Yu, Junli Huang · 2018
With the rapid development of various types of information technologies and sensor technologies, traditional industries are moving in the direction of the smart industry. The emergence of industrial Internet of Things can significantly increase manufacturing efficiency, improve product quality, and reduce product costs and resource consumption. However, the traditional signal processing based feature extraction plus classifier fault diagnosis method is no longer applicable to the characteristics of “big data” in the industrial Internet of Things. This article first proposes an intelligent diagnostic architecture for industrial IoT devices based on the edge-cloud architecture. Then based on the convolutional neural network, this paper proposes a method for fault diagnosis of equipment that is more suitable for industrial Internet of things. The method uses multi-channel to directly process the one-dimensional time series data collected by the sensor without data conversion, and the effectiveness and feasibility of the method are verified through comparative experiments.