Detection Methods of Series DC Arc Fault in Photovoltaic System Based on Three-Dimensional Feature Fusion

The present series DC arc fault (SAF) detection methods of photovoltaic system depend on complex identification methods and too simple characteristic. To address this problems, a method of SAF fault characteristic analysis and detection based on three-dimensional feature fusion is proposed. Firstly, the experimental platform is built to collect the original normal current and DC arc fault signals. Thereafter the change rate of the mean value of the current and the difference between the maximum and minimum values of the current cycle are extracted as the characteristic in the time domain. And the time-frequency characteristic is extracted by variational mode decomposition. Subsequently, the three-dimensional fault arc feature space is constructed based on the time domain, frequency domain and time-frequency domain feature vectors. The boundary parameters of the fault space are determined by experiments, and the characteristic criterion is obtained. At last, DC arc fault detection is realized based on three-dimensional feature criterion. Finally, the accuracy of the proposed method is verified by experimental analysis.