In order to predict the state of the rocket launcher electrical system, this paper proposes a data prediction method based on EMD-LSTM. Aiming at the changing characteristics of the state data of the rocket launcher electrical system, empirical mode decomposition (EMD) and long short term memory network (LSTM) are combined, and predictive analysis experiments are performed. The experimental results show that the root mean square error between the data predicted by this method and the original data is smaller, and it is better than the result of direct prediction of the original data, so it can provide reliable data for the rocket launcher electrical system failure prediction.