Multi-scale characteristic decomposition reconfigurable power line noise analysis method

The invention relates to a multi-scale characteristic decomposition reconfigurable power line noise analysis method, which aims at the field of power line noise analysis, and comprises the following steps of: after collecting power line mixed noise data, performing noise data compression based on a Gaussian random matrix, and extracting effective noise characteristic information of the noise data; and a similarity result of the power line noise database and the compressed noise characteristic data is obtained through comparison by a least square method, and then a power line noise characteristic prediction result is optimized and output by constructing an aggregated LSTM wavelet recurrent neural network, so that the prediction accuracy of the power line noise characteristics is improved. In addition, multi-scale characteristic decomposition is performed on noise characteristic data, an optimal approximation sparse solution of an estimated value and a true value of a noise signal to be reconstructed is iterati