A Similarity Measurement Algorithm for Spacecraft Telemetry Time Series

Abstract

Spacecraft telemetry time series data is the main basis for performance monitoring and real-time status analysis of spacecraft. The similarity measurement of time series is one of the important research areas. Aiming at the shortcomings of existing time series similarity measurement methods, this paper proposes a dynamic time warping algorithm based on adaptive segmentation(ASDTW). Aiming at the problem of excessive computational overhead of the Dynamic Time Warping algorithm(DTW), the algorithm divides the original sequence into several sequence segments, and defines the distance of the sequence segments according to their geometric characteristics. In the dynamic matching stage, sequence segments are used as the basic matching unit to solve the problem that the computation overhead caused by traditional point-by-point matching strategy is too high. Finally, this paper verifies the validity and feasibility of the ASDTW algorithm based on the actual telemetry data of a spacecraft. By comparing with the two baselines, the ASDTW algorithm greatly improves the efficiency of the algorithm under the premise of ensuring the measurement accuracy, and solves the problem of The problem of excessive computational time overhead of the DTW algorithm can more effectively support space mission planning.