Abstract: Due to the judgment of islanding is easy to be interfered by grid disturbance,island detection device may mal-judged islanding with grid disturbance,and the wrong judgement may cause the consequence of photovoltaic out of service.The detection device must provide with the ability to distinguish between islanding and grid disturbance.A novel approach to detect and classify islanding or grid disturbance was presented based on multi-resolution singular spectrum entropy and support vector machines（SVM）.As a signal processing method after wavelet transformation,multi-resolution singular spectrum entropy combining wavelet and spectrum analysis with entropy as output can extract the intrinsic different features between islanding and grid disturbance.Simulation results indicated that the method had good performance of accuracy and stability for the same type of samples.