Research on identification method of broken chain state of scraper conveyor based on WT-CEN

In order to realize the online identification of the broken chain state of scraper conveyor, a method of identifying the broken chain state of scraper conveyor based on wavelet transform (WT) and convolutional neural elasticity network (CEN) is proposed with the scraper conveyor of Hazmak crushing station as the research object. Wavelet transform and short-time Fourier transform (STFT) are used to preprocess and memory learning of the scraper conveyor drive motor current, and it is found that the recognition accuracy of WT-CNN is higher than that of STFT-CNN after comparison; the time-frequency domain of the current preprocessed by WT is used as the sample set, and in order to solve the phenomena of gradient explosion and overfitting of the CNN model, a convolutional neural elastic network (CEN) model is proposed to improve the controllability and stability of the model as well as the feature extraction capability. The experimental results show that the recognition accuracy of WT-CEN model for the normal operation state of scraper conveyor is 97.27%, and the recognition accuracy of broken chain state is 97.85%, which is 1.15% and 1.39% higher than that of WT-CNN respectively.