Quantitative Detection of Remanence in Broken Wire Rope Based on Adaptive Filtering and Elman Neural Network

In recent years, non-destructive testing methods for wire ropes based on remanence have attracted industry attention. The remanence detection methods have the characteristics of light equipment, high lifting value, high detection precision and low requirements on site conditions. An adaptive filtering algorithm based on wavelet decomposition was proposed to deal with the noise reduction of broken wire rope remanence data. The digital image processing method was used to locate and segment the defect. The texture features, morphological features and seventh-order invariant moments of the defect image were extracted as feature vectors, and an Elman neural network was designed to quantitatively identify the broken wires. The experimental results show that the designed filtering algorithm can effectively suppress the noise in the original signal, and the Elman recognition network has better performance of broken wire recognition.