Research on the Multi-source Measurement System of Trajectory and Data Fusion Method for Unmanned Aerial Vehicles

In the test of unmanned intelligent system, obtaining reliable extrinsic characterization information such as the location and trajectory of unmanned intelligent system accurately is the key to understanding the capability boundaries and practical application effectiveness of the system. This study focuses on the working principle, systematic error, and data fusion method of the measurement systems such as Automatic Dependent Surveillance Broadcast (ADS-B) and radar. Firstly, the working principle, data processing methods, and systematic error of the observation system are analyzed. Then, preprocessing of the measurement data if proposed, including the coordinate unification, spatiotemporal registration based on Lagrange method, extended Kalman filter algorithm. Finally, various data fusion methods are introduced, including traditional weighted fusion method, as well as modern neural network-based data fusion methods such as random forest method, K-Nearest neighbor algorithm, and long short-term memory network. This study has guiding significance for the processing of unmanned aerial vehicle trajectory measurement data.