The Research Of OMAP-L138-Based Detection System For Rotor Bar Fault In Induction Motors

Sensitively and reliably judging the rotor faults of the induction motors is becoming more and more significant. Thus, this paper is dedicated to developing a portable fault detection system of the induction motors. The research is mainly get down from two aspects, the research of sensitive and reliable detection methods and the development of rotor fault detection equipment.In the aspect of detection algorithm, there are two difficulties in detecting the BRB of the induction motors. The frequency resolution cannot be high enough and the external factors cannot be eliminated completely. In the terms of the two difficulties, the high frequency resolution spectrum estimation algorithms(multiple signal classification and extended Prony algorithm) and the singular value decomposition technique are introduced. Firstly, combining the singular value decomposition(SVD) with the multiple signal classification(MUSIC) and extended Prony algorithm introduces a new method to detect the rotor faults. Using the SVD technique to perform the stator current of the induction motors twice, the fault characteristic frequencies are clearly highlighted. The MUSIC which possesses the high frequency resolution can seek the rotor fault frequency components. And the extended Prony algorithm has a unique advantage in calculating the amplitudes.In order to simplify the detection algorithm, the Prony spectral line estimation algorithm is obtained by developing the extended Prony algorithm. Combining Prony spectral line estimation algorithm with the SVD technique we get the same excellent results. This new method gets rid of MUSIC algorithm, so the calculated quantity and the detection time of CPU are greatly reduced. Finally, combining the Park vector mode square with the estimation of signal parameters via rotational invariance technique(ESPRT) to detect the rotor bar fault can further decrease the detection time of CPU. These new methods are verified by a 3kW Y100L-2 asynchronous motor. The experimental results and the experimental data show that the new detection methods induced in this paper are sensitive and reliable.In the research field of detection equipment, through the SEED-DEC138 development board, which is based on OMAP-L138, the rotor fault detection methods of the induction motors are transplanted from PC to DSP. With the aid of the Linux system and the ARM, a set of terminal display interface is completed., which is suitable for the fault detection of induction motors.