Aiming to Complex Power Quality Disturbances: A Novel Decomposition and Detection Framework
Kunzhi Zhu, Zhaosheng Teng, Wei Qiu, Alessandro Mingotti, Qiu Tang, Wenxuan Yao
Abstract
In recent years, owing to the penetration of renewable energy and the widespread use of power electronic equipment, power quality disturbances (PQDs) have become more complex and hazardous. As the premise of power quality control, complex PQDs require more accurate and efficient detection. To address this issue, this article proposes a novel automatic method for detecting complex PQDs based on integrated intrinsic variable time-scale decomposition (I-IVTD) and weighted recurrent layer aggregation (WRLA) network. The proposed I-IVTD method reduces aliasing and endpoint effects and improves antinoise performance by innovative use of variable time scales and multiple integrations. The improved WRLA network enhances learning ability and accelerates convergence by adding three weights to each unit. The proposed framework can effectively detect 27 complex disturbances automatically and does not require manual feature design. Finally, a large number of experiments are conducted, including simulation experiments and tests on a PQD analysis platform. The test results based on the analysis platform indicate that the accuracy for complex disturbances is higher than 98%, which demonstrates the superior performance of the proposed framework. Notably, it is effective for detecting nonlinear disturbances as well.