Adaptive Neural Control for Switching Power Supplies Using Gaussian Wavelet Networks

The switching power supplies can convert one level of electrical voltage into another level by switching action. This paper proposes an adaptive neural control system for the switching power supplies. In the ANC control system, a neural controller is the main controller used to mimic an ideal controller and a compensated controller is designed to recover the residual of the approximation error. In this study, an on-line adaptive law with a variable optimal learning-rate is derived based on the Lyapunov stability theorem, so that not only the stability of the system can be guaranteed but also the convergence of controller parameters can be speeded up. Experimental results show that the proposed ANC controller can achieve favorable regulation performance for the switching power supply even under input voltage and load resistance variations.