Abstract

A load location identification method of fiber optic smart structures based on a prototype data acquisition system is proposed. It employs a genetic algorithm–support vector regression to estimate the X, Y coordinates of loads loading on a composite panel. The panel is embedded with 8 sensors. When loads act on any position of smart composite structures, the output values of embedded fiber near the loading position will change. The 8 changed signals are the features of the genetic algorithm–support vector regression. The data acquisition center collects the features as the data samples. Data samples are divided into a training set, a validation set, and a testing set. Then a support vector regression model is established by using the training set and validation set, and the parameters are optimized by the genetic algorithm. The study compares the prediction accuracy between the genetic algorithm–support vector regression model and the back propagation model. The comparison results show that the prediction accuracy of the testing set was 85.78% and it is better than the back propagation neural network prediction model. This paper demonstrates that using the genetic algorithm–support vector regression model to identify loads is not only stable and feasible but also has high precision. The presented method in this paper is significant for the health monitoring and damage identification of composite materials in the future.

© 2017 Optical Society of America

PDF Article

References

You do not have subscription access to this journal. Citation lists with outbound citation links are available to subscribers only. You may subscribe either as an OSA member, or as an authorized user of your institution.

Contact your librarian or system administrator
or
Login to access OSA Member Subscription

Cited By

You do not have subscription access to this journal. Cited by links are available to subscribers only. You may subscribe either as an OSA member, or as an authorized user of your institution.

Contact your librarian or system administrator
or
Login to access OSA Member Subscription