The fault detection algorithm of a grid-connected photovoltaic system using wavelet transform is suggested in this paper. When the faults occur in the power conditioning system, the impact on the grid system is very risky. Therefore, it is necessary to detect faults in a short time period. When using the conventional detection method, extra hardware and sensors are required to detect the inverter failure; moreover, the disadvantage of the conventional method are its high cost and re-design problem if the inverter specification needs to be changed. Multi-level decomposition wavelet transformation is an efficient method to detect the fault location and components of the inverter. Prompt and accurate diagnostic function is possible using the normalized standard deviation of the wavelet coefficients. The proposed algorithm has simple calculation and precise diagnostic capabilities of the fault detection. A computer simulation is performed and the experimental result verifies the validity of the proposed method. (C) 2016 Elsevier Ltd. All rights reserved.