Online Aerial Target Trajectory Prediction Using Multiple Model-Based Filter

Because online aerial target trajectory prediction is crucial in quadrotor applications, this paper proposes a novel online trajectory prediction method. Specifically, two motion models are firstly formulated to estimate the motion state of the aerial target, then a multiple model-based filtering scheme is designed for the target to coalesce the prediction results from the two motion models. Compared with existing prediction methods, our method not only reduces the computation burden but also achieves more accurate and swifter prognostication of the target trajectory. The accuracy of the proposed prediction method is evaluated by both numerical simulations and real-world experiments, demonstrating the superior performance over the benchmark methods.