Machine learning-enabled triboelectric nanogenerator for self-poweredcondition monitoring and regulation of USV-ROV umbilical cable

Yan Yang, Xingjia JiangHengxu Du, Fangyang Dong, Mengwei Wu, Qiang ZhaoTaili Du, Minyi Xu; Sensors and Actuators: A. Physical.

Abstract

The umbilical cable is a critical component of the Unmanned Surface Vehicde-Remotely Operated Vehicde (USV-ROV) collaborative system, which faces damage caused by stretching during operations. In this work, a tribo-electric sensor (U-TENG), composed of flexible silicone rubber and spiral electrodes, for in-situ umbilical cablecondition monitoring is proposed for the first time. Then, an integrated self-powered cable condition monitoringand regulation system is constructed to recognize different signal for different working condition based onmachine learning-enabled U-TENG. Even more, the actuator controls the winch bidirectionally to regulate cabletension and provide dosed-loop responses to abnormal conditions. Experimental results demonstrate that the U-TENG maintains stable electrical output across a strain range of 0-100 % and excitation frequencies of0.5-4.0 Hz, while also demonstrating high sensitivity to bending deformation. Cable status monitoring accuracyachieves 97.18 % under 16 operating conditions. In addition, the proposed system demonstrates excellentresponsiveness in underwater environment, preventing damage to the umbilical cable due to overload effec-tively. Thus, the proposed triboelectric sensor and control system hold significant potential for realizing ROVumbilical cable condition sensing and enhancing the safety of USV-ROV operational platforms.