基于BP神经网络PID控制器的Buck电路电压控制研究  

Research on Voltage Control of Buck Circuit Based on BP Neural Network PID Controller

作　　者：石宇扬 胡磊 肖刚 肖朋[1] SHI Yu-yang;HU Lei;XIAO Gang;XIAO Peng(School of Automation,Wuhan University of Technology,Wuhan 430070,China;Kunming Branch of Xi'an Precision Machinery Research Institute,Kunming 650000,China;Wuhan Digital Engineering Institute,Wuhan 430074,China)

机构地区：[1]武汉理工大学自动化学院,武汉430070 [2]西安精密机械研究所昆明分部,昆明650000 [3]武汉数字工程研究所,武汉430074

出　　处：《武汉理工大学学报》2024年第12期135-141,共7页Journal of Wuhan University of Technology

摘　　要：针对传统PID控制的Buck电路动稳态性能较差等问题,提出一种基于BP神经网络PID控制器在线优化的Buck电路电压控制方法,将Buck电路的输出电压、参考电压以及其差值作为神经网络输入,PID控制器参数Kp、Ki、Kd作为神经网络输出,利用BP神经网络的在线优化功能,提高控制效果。在Simulink仿真平台搭建基于BP神经网络PID控制器的Buck电路电压控制模型,进行仿真研究。仿真结果表明,文中提出的BP神经网络PID控制器超调量、上升时间等动稳态性能明显改善,参考电压突变时Buck电路电压波动更小,恢复稳态时间更短。Aiming at the poor dynamic and steady performance of traditional PID control Buck circuit, a voltage control method based on online optimization of BP neural network PID controller is proposed.The output voltage, reference voltage and their difference of Buck circuit are used as neural network input variables, and PID controller parameters K_p、K_i、K_d are used as neural network output variables.The online optimization function of BP neural network is used to improve the control effect.The voltage control model of Buck circuit based on BP neural network PID controller is built on Simulink simulation platform, and the simulation research is carried out.The simulation results show that the dynamic and steady-state performance of the BP neural network PID controller proposed in this paper is obviously improved, and the voltage fluctuation of Buck circuit is smaller and the time to restore steady-state is shorter when the reference voltage suddenly changes.

关 键 词：BP神经网络 电压控制 BUCK电路 

分 类 号：TP13[自动化与计算机技术—控制理论与控制工程] TN303[自动化与计算机技术—控制科学与工程]