Abstract: A slide mode control－common model control（SMC－CMC） controller was proposed to deal with the time－variable and complexity of the switching mode power converters．Using the state－space average modeling method，a nonlinear model of BUCK converter working on current conduction mode （CCM） was set up．Dynamic neural network which was constructed based
on a number of integrators and RBF neural network was employed to approximate the inverse model of the dynamic system． Using this inverse model to build a CMC controller．The introduction of slide mode control was to improve the robustness of controller．The modeling and simulation of system was implemented under Matlab／Simulink．The result of simulation confirms
show that the SMC－CMC control schemes are capable of reduced steady state error and improve controll‐er's reliability during supply change and load variations．

Key words: Common model control , Slide mode control , Buck converter , DC／DC converter , Non－linear control

摘要： 针对BUCK变换器的复杂性和时变性,提出了采用通用控制器和滑模-通用控制器的控制策略。应用状态空间平均建模法,建立电流连续型(CCM)Buck变换器的仿射非线性模型。利用RBF神经网络和积分器组成动态神经网络结构逼近BUCK变换器的逆模型并由此组成CMC控制器。设计了滑模-通用控制器,进一步改善了系统的鲁棒性。在Matlab/Simulink下进行了系统建模和仿真研究,结果表明该方法在输入电压和负载电阻变化时,能有效减小系统状态误差和提高系统稳定性。

关键词: 通用控制器 , 滑模控制 , BUCK 变换器 , DC／DC 变换器 , 非线性控制