Research on the application of the neural network in optical fiber temperature sensor probe design used in medical treatment

Abstract

Back Propagation (BP) and Radial Base Function (RBF) neural network methods have been used to design optical fiber temperature sensor probe which is used in medical treatment. Data trained for neural network are gained from experiment and interpolation. New kind of fiber optic temperature sensor with the best sensitivity was designed and made by the method. The selected design scheme is proved to be feasible in the experiment. This method has the characteristics of accuracy, credibility and knowledge-aid design, which are identified by the experiment. This way can save the device cost of design and decrease period of design, which has a good prospect of investigation and application.