Improved fuzzy Luenberger observer-based fault detection for BLDC motor

Abstract

Fault detection and diagnosis (FDD) are important engineering tasks that could effectively improve safety and reliability of technical processes by reducing the number of shutdowns that are necessary for traditional maintenance routines. Many techniques have been proposed in the area of FDD and the use of observers in fault detection is well established. This paper presents an improvement for Luenberger observer using fuzzy logic for detecting sensor faults. The proposed observer system is characterized by its simple design where the location of the poles of the observer is determined through exploiting a fuzzy logic approach and the gain could be adapted according to the system sates. Extensive experiments are conducted in order to investigate the effectiveness of the proposed approach. Several faults categories ranging from simple faults to complex faults are employed. The experimental results demonstrate that the improved fuzzy Luenberger observer is more effective for fault detection purposes when compared with the classical design approach for Luenberger observer.