The Control Performance Analysis of the Incremental Nonlinear Dynamic Inverse Method and Flight Test

Abstract

The feedback linearization is an effective approach to solve control problems of nonlinear systems. However, the model dependence is a critical factor to affect the control performance which is also affected by sensor noise and wind disturbance. Although the Incremental Nonlinear Dynamic Inversion (INDI) is proposed as an effective improvement based on the Nonlinear Dynamic Inversion (NDI) method to increase the model independence and robustness, the accepted scope of model uncertainty of INDI in real-world flight are rarely discussed in the literature, which is to be discussed in detail in this paper. Firstly, dominate control effectiveness (CE) parameters are analyzed focusing on the attitude control of general fixed-wing Unmanned Aerial Vehicles (UAVs). Secondly, simulation experiments are carried out to investigate the control performance in the presence of CE errors and trim changes, compared with traditional Proportion Integral Derivative (PID) method. Finally, typical actual flight experiments including dominate CE and wide speed domain tests are presented to further verify the control performance of the INDI approach. The former results show that 40% deviation from the actual dominate CE values can be adapted in the INDI method, which is important for the method to be used widely. The latter show the better adaptability of INDI in wide speed domain.