With the increasing number of nonlinear devices in the power system

the problem of harmonics is more serious and the monitoring of the power harmonics is becoming more and more important.The power quality monitoring system is increasingly mature

and comprehensive monitoring data is accumulated.In this paper

a kind of forecasting method of the harmonic monitoring data based on the deep learning algorithm of long short term memory network is proposed.The method mainly includes such four steps as harmonic monitoring data preprocessing

prediction model fitting training

prediction model application and model prediction effect evaluation.The method is realized by programming based on Python algorithm package and the actual harmonic data is predicted and analyzed.It is shown by the result that the proposed method has good prediction effect for different kinds of harmonic monitoring data

which verifies the effectiveness and practicability of the method.