Fault Diagnosis of MEMS Lateral Comb Resonators Using Multiple-Model Adaptive Estimators

In this brief a fault diagnostic unit is developed for microelectromechanical systems (MEMS) by means of multiple model adaptive estimation technique. Fault modeling tools such as contamination and reliability analysis of microelectromechanical layout enabled interpretation of microsystems behavior by evaluating their structural variations and modeling them in form of electric circuits. This technique cannot directly diagnose the faults during operation of microsystems. However, these fault-representing models can be used in multiple model adaptive estimation technique to form fault diagnosis units. Here, fault-representing systems are modeled by Kalman filters in real-time applications and are used to evaluate the fault in microsystems. MEMS lateral comb resonators are fabricated to experimentally demonstrate the fault diagnosis performance in multiple model adaptive estimation technique.