基于压缩感知和卷积神经网络的谐波检测方法研究

Study on Harmonic Detection Method Based on Compressed Sensing and Convolutional Neural Network

关键词:  造纸工业  谐波检测  压缩感知  卷积神经网络

Key Words:papermaking industry  harmonic detection  compression sensing  convolutional neural network

摘要:本文提出一种基于压缩感知（CS）和卷积神经网络（CNN）的谐波检测新方法。在CS理论下建立了具有压缩采样和重构功能的谐波检测框架，并利用CNN理论设计了具有免变换字典的重构网络。研究结果表明，本文所提出的方法具有可行性，可为造纸工业中的谐波检测提供了一种新手段。

Abstract:A new harmonic detection method based on compressed sensing （CS） and convolutional neural network （CNN） was proposed in this paper. A harmonic detection framework with compressed sampling and reconstruction function was established based on CS theory， and a reconstruction network with transform free dictionary was designed via CNN theory. The results showed that the method proposed in this paper was feasible and provided a new method for harmonic detection in papermaking industry.