Adaptive denoising for remote photoacoustic flaw detection in pillar insulators via triboelectric nanogenerator-based vibration compensation
Xingyu Chen, Kui Xu, Chaozhu Hu, Zhongju Yu, Xi Fang, Zongxuan Shi, Dingliang Wu, Baicheng Liao, Yang Yang, Zhongtao Li, Longqiang Wang, Changxin LiuPillar insulators are critical load-bearing components in power systems, and internal defects within these structures may lead to severe failures such as fracture or flashover, posing a serious threat to grid safety. Consequently, periodic and reliable nondestructive testing (NDT) of pillar insulators is important. Although photoacoustic pressure-based NDT has demonstrated excellent performance under laboratory conditions, its application in outdoor environments is severely constrained by strong background disturbances, particularly wind-induced and mechanically induced vibrations. In this study, an adaptive noise reduction strategy for photoacoustic pressure-based inspection is proposed by introducing vibration compensation using a triboelectric nanogenerator (TENG). The TENG is employed as a self-powered vibration reference source, enabling the construction of an adaptive noise cancellation framework in which the TENG output serves as the reference input. By establishing adaptive filtering models based on the least mean square algorithms, vibration-related noise components embedded in the raw photoacoustic pressure signals are accurately identified and actively suppressed. Experimental results demonstrate that the proposed method significantly improves the signal-to-noise ratio of the acquired signals and markedly enhances the resolution of the early-warning system with respect to vibration velocity. Benefiting from the purified photoacoustic pressure signals after adaptive noise reduction, the identification accuracy for internal defects in pillar insulators, such as microcracks and stress concentration regions, reaches 98.3%. Meanwhile, both false-alarm and missed-detection rates are effectively reduced. These results indicate that the proposed TENG-assisted adaptive noise reduction approach provides an effective and reliable solution for high-confidence nondestructive inspection of pillar insulators under complex outdoor environments.