Deep Learning of the SSL Luminaire Spectral Power Distribution under Multiple Degradation Mechanisms by Hybrid kNN algorithm

Abstract

The accurate prediction of the LED’s spectral power distribution under multiple degradations is essential for the lumen depreciation and color shifting. In our previous study, a gated network has been proposed to capture the SPD characteristics [1]. However, the training of such a model and be independent upon the initial guesses, a considerable learning effect is expected.In this paper, we apply the nonparametric modeling techniques, such as the k-th nearest neighborhood (kNN) method with the Fnn enhancement, and compare its prediction capability with the gate neural network. An average SPD prediction error of approximately 3-5% is observed, with 30 times shorter learning time, comparing to the pure neural network approach