Prediction of Satellite Time Series Data Based on Long Short Term Memory-Autoregressive Integrated Moving Average Model (LSTM-ARIMA)

Abstract

Time series data analysis is a method of predicting future values by observing historical data and exploring its random laws. The satellite's on-orbit operation generates a large amount of telemetry variable time series data. Satellite system state prediction with generated data plays an important role in satellite health management. However, the traditional Autoregressive Integrated Moving Average model (ARIMA) for prediction has difficulties in high precision prediction with complex inputs. Towards this aim, we propose the LSTM-ARIMA algorithm to predict the time series data of a meteorological satellite telemetry parameter and analyze the error of the prediction data. Long Short Term Memory (LSTM) neural network is more flexible than ARIMA algorithm and has room for optimization. By combining the two algorithm models by weight, LSTM-ARIMA algorithm yields high accuracy and strong reliability prediction results and mines the loss rule of the satellite telemetry parameters.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 29 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.