Texture feature extraction and its application in fault signal recognition

Aiming at the online fault diagnoses, the texture features which are usually used in image processing are firstly applied in the early fault signal recognition problems. After the parameter R based on gray-level co-occurrence matrix is defined, the parameter R extraction method of texture features is presented. Then, the novel fault signal recognition algorithm based on the parameter R of the texture feature is proposed. According to the algorithm, the pattern recognitions of the power cable in the normal state, the fault states of the single-phase open circuit, the single-phase short circuit grounding, and the two-phase short circuit grounding, and three-phase short circuit can be achieved correctly and effectively, which are proved by the simulation experiments.