Rail Crack Detection Using Optimal Local Mean Decomposition and Cepstral Information Coefficient Based on Electromagnetic Acoustic Emission Technology
Yongqi Chang, Xin Zhang, Yi Shen, Shuzhi Song, Qinghua Song, Jiazhong Cui, Huamin Jie, Zhenyu Zhao
Abstract
Rail crack detection is an essential role in the safety assurance of railway transportation. However, conventional crack detection methodologies suffer from the interference of pronounced wheel–rail rolling noise (WRRN), thereby frequently undermining detection precision. Aiming to address this issue, a novel rail crack detection method based on electromagnetic acoustic emission (EMAE) technology is presented in this article. The proposed method leverages optimal local mean decompose (OLMD) signal reconstruction algorithm, alongside a novel detection index, called cepstral information coefficient (CIC). Designed to obviate the strong WRRN interference, the OLMD algorithm has been optimized via the empirical optimal envelope (EOE), amending inaccuracies in both mean and envelope functions. Subsequently, the original signal is reconstructed by linear superposition of the first product function (PF) component from the OLMD algorithm, enhancing the information pertaining to crack characteristics. The emergent detection index CIC derives from the fusion of the primary dimensions of the gammatone cepstral coefficients (GTCCs) employing a linear transformation matrix, demonstrating exceptional proficiency in crack detection. Finally, the effectiveness and advantages of the proposed method have been demonstrated experimentally.