Optical signal parameter real-time monitoring method and device based on neural network

Abstract

The invention discloses an optical signal parameter real-time monitoring method and device based on a neural network. The method comprises the following steps: acquiring a decoupling signal after light splitting of a coupler; performing statistical processing on the decoupling signal to obtain a statistical graph of the decoupling signal; and inputting the statistical graph into a trained multi-task neural network model to obtain a real-time monitoring result of the plurality of parameters of the optical signal. According to the embodiment of the invention, no bandwidth limitation exists when the decoupling signal is subjected to statistical processing, the method can be applied to a low-bandwidth photoelectric detector with low price, and accurate real-time monitoring results of a plurality of parameters of the optical signal are obtained through a multi-task neural network model. 本发明公开了一种基于神经网络的光信号参数实时监测方法、装置，所述方法包括：采集耦合器分光后的解耦合信号；对所述解耦合信号进行统计处理，得到所述解耦合信号的统计图；将所述统计图输入至已训练的多任务神经网络模型，得到光信号若干参数的实时监测结果。本发明实施例中对所述解耦合信号进行统计处理时没有带宽限制，可以应用于价格便宜的低带宽光电探测器上，并通过多任务神经网络模型得到光信号若干参数的准确实时监测结果。