Fuzzy clustering based nonlinear system identification and controller development of Pixhawk based quadcopter

Abstract

Recently, the quadcopter configuration is becoming prevalent for both civilian and military applications, and research is continuing to improve its controllability. However, most the control methodologies depend on accurate system models which are derived from six degrees of freedom nonlinear system dynamics and they generally do not consider realistic outdoor perturbations and uncertainties. As a solution, model-free data-driven system identification and controlling of quadcopter have been refined in this paper. This nonlinear multiple input-multiple output system identification is proliferated using a fuzzy clustering model. The accuracy of the identification is seen to very high. The results of a PID controller used in conjunction with the identified fuzzy model is also presented. The PID controller with the fuzzy identified model is seen to precisely follow the commanded signals.