Spectral emissivity estimation based on K — means clustering RBF neural network

Abstract

A K-means clustering RBF neural network modeling method is introduced in this paper, this model is for infrared target spectral emissivity estimation. Part of the transmission of infrared radiation in the atmosphere is absorbed by the atmosphere, the use of RBF neural network measurement samples for analysis and learning. An infrared energy model of 3–14 μ m was established to estimate the spectral emissivity of the target at different wavelengths. The experiment results show that the maximum relative error is less than 1% compared with the theoretical emissivity calculated by the RBF network. Finally, that method is a good way to learn the spectral emissivity and it is verified by the aerospace aluminum alloy.