A nonlinear dynamic inversion-based neurocontroller for unmanned combat aerial vehicles during aerial refuelling

摘要: The paper presents the development of modelling and control strategies for a six-degree-of-freedom, unmanned combat aerial vehicle with inclusion centre gravity position travel during straight-leg part an in-flight refuelling manoeuvre. is found to have parabolic variation increasing mass aircraft. A nonlinear dynamic inversion-based neurocontroller designed process under investigation. Three radial basis function neural networks are exploited in order invert dynamics system, one each channel. Modal time-domain analysis results show that properties aircraft strongly influenced refuelling. effectiveness proposed law demonstrated through use simulation F-16 longitudinal provided interesting results, performed better than baseline inversion controller without network. On other hand, lateral-directional did not perform well as controller. It was concluded could be applied