| As the global energy supply is facing severe constraint,the photovoltaic industry has been developing rapidly in the world and been utilized widely in various fields.Hot spotting caused by partial shading or the internal factors of solar cell and series DC arc fault created by insulation aging or loose connections may easily lead to accidents,which pose a threat to photovoltaic system.Based on the analysis of the principle and characteristics of hot spot and series DC arc fault in PV system,the fault protection technologies are investigated.The modeling methods of PV modules under shading condition are studied based on analysis of the forming mechanism of hot spotting.Therefore,a simulation model of PV modules is built by using PSCAD in order to simulate the output I-V characteristics under hot spot.Consequently,a detection method of hot spot fault is introduced according to the differential of output current between shading and normal conditions,which represent as a function of the area and intensity of shadow.Based on the analysis of features of DC arcing fault in PV system,a test platform is built in accordance with UL1699 B and a series of current and voltage data are acquired.Then,the standard deviation of difference of any two contiguous time windows is calculated as the specific feature in time domain,while the sum of harmonic energy in any time window based on FFT transform is taken as the specific feature in frequency domain.Accordingly,the feature in frequency domain is processed by the Binary Gaussian Process Classifier to distinguish DC arc faults.Finally,the feasibility of the detection method is verified by experiment.The results also show that the presented method has a better immunity of electro-magnetic interference. |