Time-series Trend Analysis And Its Application In Preventive Maintenance

Mostly modern preventive maintenance based on RCM （Reliability CenteredMaintenance）analysis method or simply cycle maintenance. Domestic enterprises areoften difficult to get the probability of equipment failure when using RCM analysismethod, they usually use outside database. And plan maintenance is easy getting excessmaintenance. Therefore, this paper designed dynamic preventive maintenancemanagement system, focus on the operation trends of each equipment, using logicaloperations with other conditions, to determine whether the equipment need to preventivemaintenance.The paper introduced the research status of trends analysis algorithms and preventivemaintenance. Usingsimplelinearregressionalgorithm,AR model,Sen’s slopeestimationand Mann-Kendall to analyze the trend of the equipment operation. To choose whichalgorithm is suit for this system The result is that the Sen’s slope estimation and Mann-Kendall algorithm are suit for the system, Sen’s slope estimation is used to calculate thechange rate of amplitude, Mann-Kendall algorithm is used to analyze trend, to analyzemutation, to analyze saliency. Finally, finish the dynamic preventive maintenancemanagement system, and apply Sen’s slope estimation and Mann-Kendall algorithm tothis system. Practical application shows that Sen’s slope estimation and Mann-Kendallalgorithm have well results when to analyze equipment running trend.