Unknown Inputs PI observer-based sensor fault detection technique for BLDC motor

Abstract

Electric motors have become an essential part of the most Mechatronics system. Long operating life, higher power density and high efficiency are the advantages that make the Brushless DC motors (BLDC) widely used in industrial mechatronics applications. Downtime, costly maintenance and energy wasted are damages resulting from faults. Fault Detection (FD) techniques could help at avoiding these damages, also, they could help at increasing system reliability and operational safety. The steady state accuracy is achieved using integral term as well-known from the classic control theory. So, this work presents robust design of unknown inputs (UI) propositional-integral (PI) observer to estimate the state of the system with UI and detect the existing fault. Experimental results demonstrate the effectiveness and feasibility of the proposed technique at detecting popular sensor faults in BLDC motor.