Power quality compression method based on improved observation matrix

At present, in the research on the use of compressed sensing technology to achieve high-efficiency compression of power quality signals, the observation matrix mostly uses random Gaussian matrices. Although the random observation matrix has good non-correlation with most sparse transform bases. However, there are problems such as difficulty in hardware generation and storage, and the risk of compression failure. To solve the above problems, this paper designs a Toplitz block observation matrix based on bipolar Logistic chaotic sequence, and uses the adaptive sparsity algorithm matching pursuit algorithm (SAMP) to reconstruct the data. The simulation results show that the observation matrix designed in this paper has obvious advantages in the extraction and reconstruction of the power quality signal characteristics, and the error between the reconstructed data and the original data is small, which can better recover the original appearance of power quality signal.