基于BP神经网络的Buck变换器设计

Design of a Buck Converter Based on BP Neural Network

设计了一款基于BP神经网络的Buck变换器,神经网络控制器能根据检测到的输入电压和负载大小动态地调节PWM占空比大小。仿真结果表明,在直流电源电压加入波动范围±80 V的高斯白噪声干扰源和负载突变时,变换器输出电压稳定,整个控制系统的动态响应快,具有较强的鲁棒性和抗干扰能力。

This paper designed a buck converter based on BP neural network control. The controller can change the PWM duty cycle dynamically according to the input voltage and load of the circuit. The simulation indicates that the converter output voltage is steady while the source is interrupted by white noise ranging between＋ 80 V and- 80 V and the load is changed suddenly. The whole control system has fast dynamical response,strong robust and good anti-interruption ability.