Extraction of the Acoustic Signal from Laser Welding in a Noisy Environment with a Microphone Array
S-S. Ao, Z. Luo, Me-N. Feng and Y-F. Yan
In laser welding the acoustic signal from laser beam–material interaction is a good indicator of weld quality; however, the acoustic signal can be disturbed by other sources in a noisy environment. Generally, researchers use a single microphone to collect the acoustic signals. Owing to interference in a noisy working environment, the quality of an acoustic signal acquired by a single microphone is not good. This limits the use of an acoustic signal to detect weld quality in industry applications. In this paper, microphone array technology is adopted and an independent component analysis (ICA) algorithm is proposed. Welding experiments were conducted to verify the feasibility of proposed ICA method. To identify the extracted signals, a number of signal processing techniques, including frequency-domain and time–frequency analyses, were employed to analyse extracted signal components. The results show that typical welding defects can be identified using the proposed ICA method in noisy environments.
Keywords: Nd:YAG laser, Laser welding, acoustic signal, independent component analysis (ICA), time-frequency analyses, laser beam–material interaction