The invention discloses a household photovoltaic power station fault monitoring and early warning method. The method comprises the following steps: collecting power generation data of a user power station at regular time; comparing the real-time data with the power generation information received by the power station of the user on that day and on that month, and judging whether the power generation data is abnormal or not; establishing a power station power generation model and a power station group in the same group, judging whether the abnormal data is a power generation fault, and if so, performing fault processing; establishing a fault pool, comparing the fault information with the existing fault types stored in the fault pool, and if the comparison is successful, carrying out intelligent fault processing; if the comparison fails, starting a manual fault processing flow; after the fault processing is finished, recording the fault information, and inputting the fault information into a fault pool; and carrying out the next round of data acquisition. According to the invention, remote monitoring of the photovoltaic power station can be realized, and remote control of the fault point of the household photovoltaic power station can also be realized, so that the fault point can be cut off in time when a fault is found, and the property loss of a user is reduced.