CFRP drive shaft damage identification and localization based on FBG sensing network and GWO-BP neural networks

Carbon Fiber Reinforced Plastic (CFRP) has the advantages of light weight, high strength, small coefficient of thermal expansion, good vibration damping performance, etc., the use of its production drive shaft can reduce weight, improve the intrinsic frequency, transmission efficiency and precision, and has been used in helicopters, heavy duty machine tools, wind turbines, ships and other high-end equipment. Once the damage occurs, it will cause adverse effects such as reduction of mechanical efficiency of the system, degradation of transmission system performance, and decrease of reliability and safety. The existing non-destructive testing technology for carbon fiber reinforced plastic has limitations such as inconvenient use, complex equipment, and no real-time online monitoring. To address this problem, this paper adopts Fiber Bragg Gratings (FBG) sensors to obtain real-time strain field data of CFRP drive shaft, and uses the linear and nonlinear mapping ability of Back Propagation Neural Network (BP neural network) to identify and locate the damage of CFRP drive shaft. At the same time, the GWO optimization algorithm is used to optimize the BP neural network to construct a CFRP drive shaft damage identification and localization system based on FBG sensing network and GWO-BP neural network. The experiments show that the CFRP drive shaft damage identification and localization system constructed in this paper can obtain the CFRP drive shaft strain field information in real time and accurately identify the damage and its location.