A Health Management System Design for Large-Scale Vertical Mill

Abstract

The large-scale vertical mill is key equipment for slag and cinder being grinded process, and the health management system is a very important guarantee for the safe service and saving costs. As vertical mill has complex structure and reservoirs of state-related intrinsic data, a database based on E-R model from the requirements analysis is first constructed to manage the intrinsic data and expert diagnosis knowledge base. Then, according to the fault diagnosis inference process of experts in the field, the fault tree is derived and the diagnosis rules are conducted. Furthermore, based on the improved Apriori algorithm, the diagnosis rules are mined from the cumulative historical data and implemented to the knowledge base. Finally, the state monitoring and fault diagnosis expert system for the large vertical mill is developed. The example shows that the system can alarm the abnormal state in time and further infer the reason of the fault automatically. In addition, it can provide a reasonable explanation for the failure phenomenon.