Detection of multiple power quality disturbance events for micro-smart grids with hydrogen fuel cell

Abstract

Due to some factors such as a greater need for energy and diversification of energy sources, micro-smart grids (MSGs) with hydrogen energy (HE) based fuel cells (FCs) are becoming increasingly widespread, and these systems have effects on the power grid in terms of power quality. Single and multiple power quality disturbances (PQDs) and their levels originating from the interaction and effects of peripheral units with FC-MSG should be detected using signal processing methods. The study aims to determine the error metrics of these disturbances when multiple disturbances occur in the power system with FC-MG, to analyze whether they are FC-induced, with which signal processing methods they can be analyzed, whether a filter is needed, and to determine the effectiveness of signal processing methods for different PQD situations. In the signal processing stage, wavelet synchro-squeezed transform (WSST) is used to analyze and visualize the PQD signals. Additionally, the time-frequency representation derived by WSST is compared with that derived from the continuous wavelet transform (CWT) to evaluate its ability to visually decompose the time-varying frequency components of PQD signals. Statistical error metrics, the root mean square error (RMSE), mean square error (MSE), and correlation (Corr) values, are used to evaluate the accuracy of the results. Specifically, RMSE, MSE, and Corr improved by 7.38%, 1.04%, and 47.33%, respectively, for the four mixed disturbance signals. The results show that the proposed visual analysis tool effectively detects multiple PQD signals. Determining the type and time of the disturbance performed in this study ensures timely and correct switching of relays in FC-MSG systems. Thus, MSG and FCs are less impacted by disturbance.