Condition-based maintenance management for two-stage continuous deterioration with two-dimensional inspection errors

Abstract

Inspections often perform imperfect outcomes during maintenance processes owing to human errors, management issues and other limitations. In particular, such imperfection affects the maintenance management of multistage deterioration significantly due to both false state identification and measurement errors, whose quantitative analysis, however, is seldom reported in the literature. To fill these gaps, this paper devises a condition-based maintenance management strategy oriented to two-stage continuous degradation under two-dimensional inspection imperfection. Specifically, a threshold-based replacement is executed under the normal-working state if the detected degradation value exceeds the preset limit; additionally, preventive replacement is immediately performed once the defective state is identified. Notably, the detection outcome rather than the actual working condition decides how preventive maintenance operates. The long-run cost rate is minimized via the optimization of the inspection cycle and replacement limit. Besides, numerical experiments conducted on train bogie bearing are provided, showing substantial superiorities over cost-effectiveness promotion and performance improvement.