In the article, experimental studies of the neurocontroller of a laboratory installation of a quadrocopter link are presented. The latter is a nonlinear dynamic system. The input vector of the neurocontroller included the angle of inclination of the rod (a beam of the quadcopter) relative to the horizon, its angular velocity, and the angular velocity of the propeller. The output signal of the neurocontroller is proportional to the supply voltage of the propeller drive. In the article, the planning of experimental studies was carried out and eight indicators were selected, according to which the quality of the control process was evaluated. In addition, a qualitative analysis of the control process of the dynamic system motion was carried out with the corresponding graphical dependencies. The data obtained showed a good quality of control at a zero setpoint angle. For other values of the setpoint (-0.52 and -1.05 rad), the neurocontroller provides the rod aboutness to the setpoint angle and the control stability. However, the quality of control is not high. The reason for this effect has been established in the work. In order to improve the quality of control, the neurocontroller was modified by including an integral component in its structure. At the same time, the steady-state control error has significantly decreased with minor changes in other estimated indicators.
Control Method for Nonlinear Dynamic System Using Neural Network.