Structural States Monitoring with Fiberoptic Sensing Array and Probabilistic Neural Network

Abstract

In this paper, structural states of a simple composite plate specimen are monitored using the fiberoptic sensing array and the probabilistic neural network. Firstly, structural states data were collected by a data collection system based on the fiberoptic sensing array embedded within the material, and were utilized as feature vectors of structural states. Secondly, structural states to be monitored were defined according to the load location and the load level. And the number of states was reduced by grouping neighboring elements to one category. Finally, the probabilistic neural network used for structural states monitoring was designed. The smoothing parameter in the probabilistic neural network was studied. And structural states were identified based on collected data. When this trained network is subjected to the measured response, it should be able to monitor the load location and the load level. The effectiveness of the proposed method was demonstrated.