Method and system for predicting transmitted spectrum of optical fiber surface plasmon sensor based on neural network

Abstract

The invention relates to an optical fiber surface plasmon transmitted spectrum prediction method and system based on a neural network, and particularly relates to the field of transmitted spectrum detection methods. The method comprises the following steps: acquiring gold film thicknesses and gold film widths of a large number of target optical fiber surface plasmon sensors, a refractive index of a to-be-measured environment and corresponding transmission spectrums, and normalizing the gold film thicknesses, the gold film widths and the refractive index of the to-be-measured environment; building a neural network model; inputting the normalized thickness and width of the gold film, the refractive index of the environment to be measured and the corresponding transmission spectrum into a built neural network model, and training the model to obtain a trained neural network model; and substituting a plurality of groups of gold film thicknesses and gold film widths which do not participate in training and the refractive index of the environment to be measured into the trained neural network model, and predicting the transmission spectrum of the target optical fiber surface plasmon sensor. The credibility of predicting the transmitted spectrum through the trained neural network model is high.

Description

technical field [0001] This application relates to the field of transmission spectrum detection methods, and specifically sets up a method and system for predicting fiber surface plasmon transmission spectrum based on neural networks. Background technique [0002] With the development of technology, machine learning has become a hot scientific research direction. Machine learning is a science of learning from data through programming. Neural network is a branch of machine learning, which can deal with classification problems and regression problems. Classification problems such as: identifying whether a picture is a cat or a dog; regression problems such as: house price prediction, light transmittance prediction. We can regard the neural network as the human brain. Through repeated learning and memorization of a certain subject, the human brain naturally responds to relevant formula diagrams whenever a certain knowledge point of the subject is mentioned. The same is true of...

AI Technical Summary

Problems solved by technology

[0005] The object of the present invention is to provide a fiber surface plasmon transmission spectrum prediction method and system based on a fully connected layer neural network model to solve the problem of detecting the surface of an optical fiber every time in the prior art. The transmission spectrum of the plasmon sensor needs to build an optical scene, which makes the detection process of the transmission spectrum of the optical fiber surface plasmon sensor more complicated, and it is easy to introduce errors during the detection due to human reasons

Images