I- Introduction

II- Quaternion

III- Statistics

IV- Classification Tree

V- Algorithm

References

Abstract

Induction motor fault identification is essential to improve efficiency in industrial processes improving costs, production line and maintenance time. This paper presents a novel motor fault detection methodology based on Quaternion Signal Analysis (QSA). The proposed method establishes the quaternion coefficients as the value of motor current measurement and the variables x, y and z are the measurements from a triaxial-accelerometer mounted on the induction motor chassis. The method obtains the rotation of quaternions and applies quaternion rotation statistics such as mean, cluster shades and cluster prominence in order to get their features, and these are used to classify the motor state using the proposed tree classification algorithm. This methodology is validated experimentally and compared to other methods to determine the efficiency of this method for feature detection and motor fault identification and classification.

INTRODUCTION

INDUCTION motors are important in industrial processes. Motor fault detection improves costs, production line and maintenance time. The induction motor is affected by electrodynamic forces, winding insulation, large voltage stresses, thermal aging and mechanical vibrations from external or internal sources [?]. The typical mechanical faults that occur in induction motors are stator fault, bearing fault, broken bars fault and rotor fault. The importance of the evaluation of induction motors has motivated the development of algorithms for early fault detection [?]. Bearing and rotor faults are the most common type of analysis performed on induction motor failures. For instance, in [?], a stator inter-turn fault is detected using an algorithm that analyzes impedance. The authors in this paper calculate the impedance using winding function theory, and then they compare their results with those in a database to detect a fault classification. A statistical method such as the least-mean-square is used with this kind of algorithms to obtain bearing fault detection by analyzing the vibration signals that are taken throughout the endurance test [?]. Similarly, spectral kurtosis-based algorithms are used to find faults in the bearing elements by using the stator current or the vibration signal to analyze the characteristic frequencies [?]- [?].