Fault estimator and diagnosis for electric motor in coal mine via self-constructing fuzzy UKF method

Article type: Research Article

Authors: Liu, Zhiyong^{; *} | Jia, Fangyun | Wang, Ali | Luo, Lianhe

Affiliations: Xian Yang Vocational & Technical College, Xian Yang, China

Correspondence: [*] Corresponding author. Zhiyong Liu, Xian Yang Vocational & Technical College, Xian Yang 712000, China. E-mail: waljxzy@163.com.

Abstract: This study investigated fault information estimation and diagnosis using a novel approach based on an integrated fault estimator and state estimator for an electric motor in coal mine. The proposed scheme uses a self-constructing fuzzy unscented Kalman filter (UKF) system to simultaneously estimate the system state and approximate the fault information. To achieve this, a generalized linear discrete-time system of the electric motor in coal mine without faults was first transformed into an equivalent standard state-space system with faults. Then, the self-constructing fuzzy UKF system was designed in order to obtain the fault information. According to fault information obtained fault detection experiments based on fuzzy clustering were performed with the proposed scheme and the fault feature parameters required for fault isolation were determined. Finally, the scheme was applied to an electric motor in coal mine to demonstrate the effectiveness of the proposed fault estimation and diagnosis approach. Results of the simulation illustrate the effectiveness of the proposed method.

Keywords: Self-constructing fuzzy system, unscented Kalman filter (UKF), state estimation, fault information, electric motor, coal mine

DOI: 10.3233/JIFS-190755

Journal: Journal of Intelligent & Fuzzy Systems, vol. 37, no. 5, pp. 6879-6890, 2019