Synthesis of neural network control in the sliding mode of the ventilation and air conditioning system

A method for synthesizing a sliding-mode neural network regulator for controlling an air conditioning system is proposed. The regulator is synthesized on the basis of sliding mode control methods and Lyapunov functions. To increase control accuracy under varying plant parameters and in the presence of external disturbances, the gain of the regulator is adjusted using a neural network. The algorithm for learning network weights is based on the velocity gradient method. The stability of a closed-loop control system with the proposed regulator is investigated by the Lyapunov function method. To evaluate the effectiveness of the developed regulator, the article compares its characteristics with other control methods in the computing program package MatLab/Simulink. The conducted modeling makes it possible to justify the choice of the optimal control method for real systems.

Authors: M. D. Nguyen, M. P. Belov, A. M. Belov

Direction: Electrical Engineering

Keywords: ventilation and air conditioning systems, sliding-mode neural network regulator, Lyapunov function, velocity gradient method, MatLab/Simulink

---

View full article