Incremental Prediction Method of Optical Performance Degradation Trend Based on Deep Learning

Abstract

This paper proposes a deep learning-based incremental prediction method for optical performance degradation trend, constructs a time series matrix of performance indicators, and adopts a singular value decomposition method to find the high-quality core indicators affecting the optical performance degradation. The incremental index influence degree collaborative analysis method is proposed to construct the degradation matrix and incremental matrix, and the incremental matrix is used to dynamically adjust the neural network weights to optimize the prediction model of optical performance degradation.