Handling-Qualities Based System Identification and Flight Control of an Unmanned Helicopter

The main difficulties in modelling dynamics of a helicopter arise from the structure complexities and dynamic uncertainties of rotorcrafts. This paper illustrates a simple and accurate identification approach for helicopter dynamics modelling. To minimize the modeling complexities, dynamic model structures are derived from main-rotor and tail-rotor subsystems through comprehensive analysis. To cope with the modeling uncertainties, handling qualities (HQ) based parameters, instead of ordinary model parameters, are identified from frequency-sweep test data through particle swarm optimization (PSO). The identified models provide excellent match with the flight data around the hover point. PID flight controllers were designed on the basis of the established models. The desired performance of the controller implemented on the helicopter was demonstrated by the automatic flight experiments.