Lube Intelligent Diagnosis System Combining Bayesian and BP Network Based on IOT Technology

Abstract

The existing industrial lubrication depend on experience judgment, off-line inspection and regular oil change, whose maintenance requires rich personnel experience and still always have many errors. Line monitoring and quality diagnosis for industrial lube were studied to establish the distributed the online monitoring system based on hierarchical structure, information fusion diagnostic system based on Bayesian network and BP neural network. The filtering system for industrial lube has been developed to achieve unattended, automatic operation purposes, and trialed in the metallurgical industry. The results show monitoring data is stable, reliable, and the problem of high water content of lube in the steel industry is solved. At the same time, lube filtering is transformed from the traditional blind continuous filtering to real-time targeted filtering. In the premise of guaranteeing the lube quality, the system can save electricity more than 30%.