The invention discloses an automatic early warning method for power equipment faults based on infrared imaging temperature measurement, and relates to the technical field of power equipment detection and maintenance. The main method step is to take the four data sets of T, T20, Tmax, and Tmax20, with the time as the X axis and the value as the Y axis, and perform incremental recursive regression modeling respectively, and the real-time temperature measurement data sets formed by each are used as the trend The basic data of the analysis; according to the trend model of T, T20, Tmax, Tmax20, predict the temperature value that T, T20, Tmax, Tmax20 will reach in the next 1 hour, 1 day, 1 week, according to the predicted temperature value, and alarm Threshold comparison to realize trend warning. For the first time, the present invention applies the trend analysis of the conversion temperature of infrared temperature measurement combined with environmental climate information to the judgment of infrared temperature measurement trend warning of electric power equipment; for the first time, the infrared temperature measurement combined with environmental climate information and current load to calculate the maximum duration of the equipment The temperature of the hot spot of the load is analyzed, and this index is analyzed to predict the temperature trend, and further realize accurate and advanced warning.