Research On Predictive Maintenance Model Of Equipment Based On Dynamic Bayesian Network

With the spread of the concept of industry in the world,the reliability of the equipment in the production of higher and higher requirements to ensure that the production system equipment in the normal and reliable operation of today's manufacturing enterprises to improve market competitiveness of the powerful weapons.The research on equipment maintenance management can provide strong support for the normal operation of the equipment.Now,the maintenance and management of the equipment are mainly based on the prevention and maintenance of time-based,and this method is difficult to control the maintenance cycle,which is likely to cause over-maintenance or timely maintenance.Therefore,this paper studies the predictive maintenance of state-based equipment,analyzes and extracts the characteristic variables that reflect the health status of the equipment,and monitors these characteristic variables.The dynamic Bayesian network is used to predict the health status of the equipment.Based on the prediction of the health of the equipment put forward the equipment of the predictive maintenance strategy for the reliability of equipment operation,reduce equipment maintenance costs have very important theoretical value and practical significance.Firstly,the fault tree model is established by the fault tree analysis,and the key factors affecting the operation of the equipment are obtained.Then,the failure mode and the hazard analysis of these factors are finally obtained.The characteristic variables needed to be monitored are provided for the prediction of the health status of the equipment.Based on the dynamic Bayesian network equipment health status prediction model,the monitoring of the characteristic variable value as evidence of the input dynamic Bayesian network equipment state prediction,the equipment health status change trend;Finally,the traditional maintenance Strategy to improve,taking into account the equipment recession and maintenance of the impact of equipment operation,based on the establishment of equipment health status prediction of the predictive maintenance strategy,by establishing the equipment health maintenance threshold and predictive maintenance cycle,so that the appropriate time to equipment maintenance,To avoid the past after the maintenance and cycle-based maintenance easily lead to system reliability and maintenance costs are too high problems.In this paper,we establish a predictive maintenance model based on Dynamic Bayesian network to improve the reliability of the equipment while saving the maintenance cost of the equipment.