基于BP神经网络PID控制DC/DC变换器研究  

Research on Controlling DC/DC Converter Based on BP Neural Network PID

作　　者：胡霞[1] 王华俊 HU Xia;WANG Huajun(School of Electrical and Information Engineering,Anhui University of Science and Technology,Huainan Anhui 232001,China)

机构地区：[1]安徽理工大学电气与信息工程学院,安徽淮南232001

出　　处：《佳木斯大学学报（自然科学版）》2023年第3期28-30,95,共4页Journal of Jiamusi University：Natural Science Edition

摘　　要：针对传统PID控制DC/DC变换器具有响应速度慢,输出电压纹波系数大等问题。提出了一种BP神经网络结合PID的控制方法。对神经网络的结构进行了分析,对算法进行了数学公式推导,最后采用Matlab软件编程实现BP神经网络算法。在Simulink模块中搭建了Buck电路模型,再分别搭建了传统PID控制和模糊PID以及BP神经网络控制模型对其进行仿真,得到相应的输出电压仿真结果图。仿真结果表明,BP神经网络PID控制下的输出电压更稳定,纹波系数为0.4%,小于其他两种控制方法。For traditional PID-controlled DC/DC converters,it has problems such as slow response speed and large output voltage ripple coefficient.In this paper,a control method combining BP neural network with PID is proposed.The structure of the neural network was analyzed,the mathematical formula of the algorithm was derived,and finally the BP neural network algorithm was implemented by Matlab software programming.The Buck circuit model is built in the Simulink module,and then the traditional PID control and fuzzy PID and BP neural network control models are built to simulate them,and the corresponding output voltage simulation result diagrams are obtained.The simulation results show that the output voltage under the control of BP neural network PID is more stable,and the ripple coefficient is 0.4%,which is less than that of the other two control methods.

关 键 词：BP神经网络 PID DC/DC变换器 负梯度下降 

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]