Active noise control method of power communication signal based on BP neural network

Abstract

In the power line communication network, the electrical node switch connected to the Power Line Communication network will affect the impedance, noise and signal attenuation of the network. The connection change between the network nodes will lead to the change of the topology structure. This paper proposes a new type of automatic modulation pattern recognition system assisted by blind channel recognition based on a convolutional neural network, which is composed of two convolutional neural networks with different architectures, and successfully improves the recognition accuracy of the traditional automatic modulation pattern recognition system. The Signal to Noise Ratio recognition rate of the proposed blind channel recognition assisted Adaptive Multi-Rate system is as high as 99 %, which is more than 5 % higher than that of other comparison methods. After several iterations, the accuracy and loss function of the model studied in this paper are still at a stable level. The intelligent recognition technology of blind signals proposed in this paper can effectively improve the interference problem of communication signals in power system and has a promotion effect in other communication systems, which can achieve higher technical application value.