Study of Microwave Dielectric Resonator Oscillator Frequency Degradation Model for Intelligent Maintenance Systems

This paper presents a modeling method to model and assess the Dielectric Resonator Oscillator (DRO) performance degradation to supply a basis for the fault diagnoses, health state estimates and life prediction of the DRO in the Intelligent Maintenance Systems (IMS). In our research, the model combines internal state transition model with the Brownian motion stochastic process to show the degradation mechanism and random factors respectively. The degradation models parameters were recognized by the degradation data coming from the Accelerated Degradation Test. The result indicates that the model significantly describes the DRO practical degradation process with an impressively small sum of squared residuals. We also resolved the curvilinear boundary crossing issue of Brownian motion with optimal linear approximation methods and presented an approximate life distribution.