The Research On Detection Algorithm And Device For Broken Rotor Bar Fault In Induction Motors

Detecting the broken rotor bar fault(BRB) in induction motors rapidly and accurately has important significance. This thesis addresses to this topic in both algorithm and hardware aiming to improve the practicability of the algorithm and develop a portable detection device.As for algorithm research, a new method for detecting broken rotor bar fault(BRB) in induction motors is proposed which is based on the combination of singular value decomposition(SVD) filter, speedy fourth-order cumulants(SFOC) and estimation of signal parameters via rotational invariance technique(ESPRIT). Simulations of BRB show that SVD filter can effectively filter out the fundamental component and background noise in stator current signals, and thus, highlight the BRB feature component. On the other hand, ESPRIT based on fourth-order cumulants can effectively reduce the noise interference, extend signal array, and hence, extract BRB feature component with a high frequency resolution. Especially, the combination of SVD and ESPRIT based on fourth-order cumulants shows a promising benefit on extracting BRB feature component with a high resolution even with short-time signals available. Additionally, in order to speed up ESPRIT based on fourth-order cumulants, a reduction algorithm is proposed to eliminate redundant data of the uniform linear array DOA(direction of arrival) estimation, thus greatly reducing the calculation amount of this kind of ESPRIT. As a result of this, SFOC-ESPRIT is yielded as a new type of ESPRIT. At last but not least, by combining SVD and SFOC-ESPRIT, a detection method for BRB in induction motors is presented. A series of experiments are conducted to verify the presented detection method.In terms of hardware research, this dissertation describes the hardware structure and function of SEED-DEC138 demo board, builds an embedded Linux development platform and configures the development environment for the demo board. At last, a graphical user interface(GUI) with Qt software is developed and transplanted to the demo board as a friendly man-machine interface.