Supraharmonics monitoring based on VSSESP-DBP dynamic compressed sensing algorithm

Chao Yan ¹ Hongtao Shen ¹ Jie Yu ² Peng Tao ¹ Hongxi Wang ¹ Ting Yang ^2*

• ¹ State Grid Hebei Electric Power co. ltd, Shijiazhuang, Hebei Province, China
• ² Tianjin University, Tianjin, China

With the advancement of power electronic devices toward intelligent high-frequency operation and the widespread integration of distributed renewable energy sources, issues related to electrical power quality particularly those arising from superharmonics are becoming increasingly significant. However, the non-stationary and wide-frequency characteristics of superharmonic signals pose significant challenges in their effective monitoring. To address these challenges, this paper proposes a supraharmonics monitoring approach based on the VSSESP-DBP dynamic compressed sensing algorithm. First, due to the difficulty of accurately sampling non-smooth superharmonic signals with traditional static time windows, a dynamic time window is applied to the superharmonic signals at the sampling end, and the feedback-type flexible modulation of the window width is realized by introducing a scale stretch factor, which effectively reduces the reconstruction error. Then, based on the sparsity of the superharmonic signal within the time window, its sparsity is inverted to prove the high efficiency and reasonableness of applying the compressed sensing theory to realize the dynamic compressive sampling of the superharmonic signal. Second, to overcome the drawbacks of high computational complexity and poor real-time performance of traditional algorithms in continuous reconstruction, the VSSESP-DBP dynamic reconstruction algorithm is designed at the reconstruction end, based on the variable step-size sparsity self-estimating subspace tracking (VSSESP) algorithm to find the initial solution, and further proposes a dynamic basis tracking (DBP) algorithm to exploit the time-dependence of the signal support set, and use the solution of the previous moment as the a priori information, which effectively improves the solution speed of the reconstructed signal. The experimental results show that the proposed method can realise the dynamic monitoring and reconstruction of superharmonics, which effectively reduces the amount of sampling data and further improves the accuracy and efficiency of reconstruction compared with the traditional method.