On a Self-diagnosing System for a Liquid Core Optical Fibre (LCOF) Intelligent Composite Structure
L-B Shen, Z-M. Zhao, Y-W. Wu, Y-S. Yu, K. QIAN and J-L. Liu
According to the function of self-diagnosis and self-healing for smart composite structures health monitoring system, this paper proposed a smart structure based on liquid core optical fibre (LCOF) sensor network. A model of loads recognition algorithm based on particle swarm optimization-support vector regression (PSO-SVR) is presented to achieve the self-diagnosis of the LCOF sensor system. Furthermore, the loading experiments were conducted on the E-51 composite structures plate embedded with eight liquid core fibre sensors. The relationship between the changes of light intensity in fibre sensor and loads positions was researched on. The support vector regression (SVR) model was established using sample data to predict loads position. And the parameters were optimized by particle swarm optimization (PSO) algorithm. And the test error is 0.245%. The study showed that this method can effectively determine the loads positions with small size training sample. It realized the self-diagnosis system for liquid core fibre sensors and it provides a foundation for the future research on the self repairing for smart composite structures.
Keywords: Liquid core optical fibre, smart composites structure, loads recognition, particle swarm optimization (PSO), support vector regression (SVR)