<![CDATA[Quadcopter altitude control is a critical issue in unmanned aerial vehicle systems due to its nonlinear dynamics and impact on operating conditions. Various fuzzy logic-based approaches have been used to address these nonlinear characteristics. However, in many previous studies, the rule structure and membership function were determined heuristically or based on expert knowledge, resulting in a less systematic design process that did not fully represent the distribution of system dynamics data.This study proposes a data-driven fuzzy control approach using the Gustafson–Kessel clustering algorithm to automatically generate fuzzy rules from system data. This algorithm is used to identify an ellipsoidal cluster structure in the input space formed by the error variable (e) and the error change (Delta e). The cluster center parameters are used to construct Gaussian membership functions and a rule base for the fuzzy inference system that generates control signals for the quadcopter's vertical dynamics.Evaluation is conducted through quadcopter altitude control simulations using several error metrics. The simulation results show Integral Squared Error value of 0.363789, Integral Absolute Error of 0.554409, Root Mean Square Error of 0.190637, and Mean Absolute Error of 0.055386, with a maximum error of 0.999600 in the beginning of the system response.]]>
Copyrights © 2026
