Real-time reconstruction of multi-area power system signals based on compressed sensing

With the growing complexity of the power systems, increasing quantity of signals needs to be transmitted in multi-area power networks. In the light of the traditional signal acquisition process, Shannon sampling theorem need to be contented to prevent signal distortion, so the transmission of massive information requires large bandwidth. The timely and accurate transmission of real-time signal poses great challenges to multi-area power systems. This paper puts forward a novel signal processing means for the real-time reconstruction of multi-area power system signals via compressed sensing. Compressed sensing is an effective approach for acquiring compressed signals at a sampling rate remarkably below the Nyquist rate and accurately reconstructing signals from measurement vector. Multi-area power system signals needs to be sparsely transformed using sparse basis before being compressed. This paper uses an over-complete dictionary as the sparse basis. It can better match the structure of signal and achieve better sparse representation. In addition, this paper utilizes a temporality multiple sparse Bayesian learning (TMSBL) algorithm for improved reconstruct signal. Simulation results demonstrate that the signal can be accurately reconstructed applying the advanced modus while significantly reducing the signal reconstructed time.