Analysis and application of deep space exploration satellite telemetry data combined with empirical mode decomposition and ARIMA model

Abstract

With China’s deep space exploration program constantly expanding, the number of satellites on-orbit is increasing year by year. The stable operation of satellites in orbit hinges on long-term monitoring and analysis of the telemetry data downloaded from the satellite on the ground, which enables quick detection of any irregularities. If we can grasp the changing trend and operation law of satellite telemetry data, we can monitor and control satellites and reduce losses better. Aiming at a large amount of telemetry data of deep space exploration satellites, the data presents non-stationary and nonlinear characteristics, The time series to be processed is constructed sequentially using the acquired telemetry data, and it is decomposed using the empirical mode decomposition approach. The non-stationary signal is divided into a stationary sequence and the sum of a series of stationary intrinsic mode functions. Then the decomposed sets of results are used as the input of the ARIMA model, and the ARIMA model is constructed to predict the predicted value of each decomposed sequence, and the weighted summation of the predicted values is performed to reconstruct the predicted results. Lastly, actual satellite telemetry data is used to confirm the method’s viability. The experimental findings demonstrate that the model can more successfully anticipate the short-term trend of satellite telemetry data.