This paper presents the development and simulation of an intelligent system for controlling a drone through hand gestures, evaluated and tested in the MATLAB/Simulink environment. First, a detailed mathematical model of the drone dynamics is built, including aerodynamic forces, motor dynamics and battery model. For precise estimation of the system states the model also incorporates the sensor noise, processed with an Unscented Kalman filter. Flight stabilization is achieved through a PID control system. The central contribution of the paper is an implementation of an intelligent control of a drone using a smart glove, equipped with flex sensors and an inertial measurement unit. The smart glove raw analog signals are processed and classified by a neural network with an 8–10–4 architecture, trained using the Levenberg-Mar-quardt algorithm. The network classifies the fused sensor signals into four discrete control commands, which are fed directly into the drone's flight control system, as structured reference inputs. The integrated system is verified through a simulation, where the drone successfully executes a three-dimensional trajectory, commanded exclusively by the hand gestures. The results show stable tracking of the reference signals, without oscillatory instability, confirming the practical applicability of the proposed approach.