Review on Neural Network Identification for Maneuvering UAVs

Unmanned Aerial Vehicles (UAV) need high mobility in performing military tasks. In this paper, we summarized the UAVs identification methods based on neural network, which takes the characteristics of strong nonlinearity and coupling into consideration. These methods provide some feasible approaches for identifying small maneuvering UAVs. Maneuvering UAV systems are required to be identified more quickly and accurately. In recent years, improvements in many aspects bring new development opportunities to identify maneuvering UAV systems, such as the increase of sampling rate of sensors, the development of computational fluid dynamics (CFD) and embedded technology, the widespread use of deep learning in control area and the rise of neural network chips. We are expected to get more precise maneuvering UAV models based on neural network identification methods with the help of more preeminent software and more advanced hardware.