Volume 4 Issue 1 ( March 2017 )

Pages 58-66

System Identification and Adaptive Control of Mass-Varying Quad-Rotor

Tarek N. Dief, Shigeo Yoshida

[ABSTRACT ]

Modelling, system identification and the controller for variable length quad-rotors are presented in this paper. Modelling was derived using Newton-Euler method, then small disturbance theory was used for linearization. Plackett’s algorithm was used for system identification to predict the system parameters. Mass-varying problem is the main objective of this paper, and the effect of changing the system parameters was discussed in detail. The system parameters are updated in real-time during flight with low sample time. The PID-Accelerator controller of the quad-rotor was updated also in real-time based on the change of the system identification output. The attitude and altitude control of the quad-rotor were presented using an adaptive classical controller. Now the tethered mass-varying quad-rotor can be simulated. The objective of this work is to make the quad rotors fly for long time and to be robust for variable inputs comes from the tether. The simulation results of the system identification and control responses of the attitude and altitude are presented in this paper. The disturbance of the wind was also considered in the controller design.

Keywords: System Identification, Plackett’s algorithm, Quad-Rotor, PID-A controller.