A Research Of Switching Power Supply By Dynamic Neural Network Inverse Control

In recent years, with the rapid development of information technology, a portable equipment is closely related to people’s daily life, work, which is inseparable from the reliable performance of power. Compared with linear regulated power. Switching power supplies with low power consumption, high efficiency, excellent characteristics, are widely used in many industries. As a kind of strongly nonlinear time-varying systems, it can’t be analyzed and controlled by classical control theory. Therefore, the researches of model and control method have important theoretical and engineering value. With the development of intelligent control, more and more advanced control method is applied to it, and in specific situations it has achieved good control effect. In this paper, the theory of neural network is used in switching power supply control, and combined with advanced control methods, in order to improve it’s performance. Through the Matlab/Simulink simulation, we can verify the effectiveness of the presented control method. In this paper, main work is as follows:First, dynamic neural network inverse control is introduced, based on the study of traditional small signal modeling method and circuit structure, control principle.By using feed-forward dynamic neural network to establish inverse model of current-mode switching power supply with voltage loop, combined with linear inverse system to resolve the coupling principle of dynamic neural network inverse control method is proposed to improve the performance of current mode control of switching power supply.Later, in order to improve the tracking precision of the dynamic neural network inverse control method and dynamic performance, introducing general model control theory, the dynamic neural network inverse control combined with common model control strategy, composite dynamic neural network inverse control method is proposed, improved the control performance of current-mode switching power supply.Finally, the adaptive control theory is introduced, and try to apply the adaptive method to improve the composite dynamic neural network inverse control, an improved composite dynamic neural network inverse controller is designed, which improves the robustness and dynamic performance of the controller, further improved the control performance of current-mode switching power supply.