Grey box identification approach for longitudinal and lateral dynamics of UAV

Abstract

The modeling of aerospace vehicles using system identification technique is a very effective and significant approach in today's industry. The different analytical methods cannot accurately model the dynamics of unmanned aerial vehicle (UAV). The model obtained using system identification technique represents the UAV in different flight envelopes which lead to development of effective flight control systems. The models for longitudinal and lateral dynamics of UAV is obtained using first principle approach. The recorded flight test data is processed using system identification toolbox of Matlab which result in obtaining grey box models for UAV dynamics using Prediction Error Method (PEM). Validation of both longitudinal and lateral models is carried out along with performance of error analysis. Also the different aerodynamic parameters are obtained for these models. The accuracy of validation results show that these models can be used for flight simulator, autopilot design and other different controller design. UAV model used as reference is SmartOne.