Research Of Induction Motors Fault Detection Based On EMD And ICA

Induction motors have been widely used in industrial production fields such as aviation, space flight and navigation because of its simple structure, low cost and ease of use. Its unexpected breakdown not only can bring huge economic losses but can also result in casualties or even loss of life, so it is necessary to detect anomalies of the induction motor on time.Empirical mode decomposition (EMD) is a new kind of signal processing method, which performs well in self-adaption decomposition and local analyzing. For its outstanding advantages, it is fit for analyzing non-stationary signals. Independent Component Analysis (ICA) is one of the blind source separation methods which have the capability to decompose multiple observational signals into mutual independence components. ICA has already received much attention in the signal processing field and has been used widely in feature extraction, image processing and process monitoring, etc. The purpose of this essay is to combine EMD and ICA to detect bearing fault and broken rotor bar of the induction motor.The main contents of this thesis are as the following:1, The fundamental rules or principle of the detecting bearing fault and rotor broken is analyzed. The fault mechanism of the two faults is discussed by using the fault frequency characteristics analysis. The existing fault detection and diagnosis methods are further studied in this paper.2, This paper provides an in-depth analysis and research on the theory of EMD and ICA, and demonstrates the merits of these two algorithms by using simulation results.3, A new fault detection method based on EMD and ICA is proposed. Firstly, the stator current signals are decomposed by EMD. The signal-to-noise ratio (SNR) of the current signals is enhanced by removing the components of intrinsic mode functions (IMFs) in high frequency. Then, ICA is used for extracting the dominating features from the signals. The experimental results demonstrate that the proposed method can be well applicable for detecting faults of induction motors.