Multi-area power control method based on real-time compressed sensing

TL;DR: In this paper, a multi-area power control method based on real-time compressed sensing was proposed, where original signals are sampled at a rate far smaller than a nyquist rate; obtained measurement values of which number of dimensions is smaller than the number of original signals; and the original signals were precisely recovered from the measurement values through a reconstruction algorithm.

Abstract: The invention provides a multi-area power control method based on real-time compressed sensing. According to the compressed sensing, original signals are sampled at a rate far smaller than a nyquist rate; obtained measurement values of which number of dimensions is smaller than the number of dimensions of the original signals are transmitted; and the original signals are precisely recovered from the measurement values through a reconstruction algorithm. According to the compressed sensing, through adoption of a communication network with a limited bandwidth, relatively many signals can be transmitted and the signal transmission efficiency is improved. According to the method, the compressed sensing is introduced, when the power signals are transmitted in multiple areas, the number of dimensions of the transmission signals is reduced, and the signals still can be precisely reconstructed in a reconstruction buffer. The power signals are non-sparse signals in a time domain, sparse representation needs to be carried out on the power signals. According to the method, a dictionary learning algorithm is employed; according to a sparse dictionary trained through adoption of the dictionarylearning algorithm, the power signals can be converted into sparse vectors; when the signals are reconstructed, the sparse vectors are reconstructed from the sparse dictionary and a measurement matrixthrough adoption of the reconstruction algorithm; and finally the original signals are recovered through adoption of the sparse vectors.