中文摘要：

针对传统线性PID控制对复杂控制对象难以建模及人工整定经验缺乏等问题,提出了一种双闭环Buck变换器系统的模糊PID控制策略。首先,设计了电压外环和电流内环结构以同时提高系统的抗扰性和跟随性;其次选择高斯函数和三角形函数相结合的隶属度函数以使控制器对小误差的灵敏度得以增强;接着,针对PID参数分别设计了3个子推理器并通过基于专家经验的规则库制定出Buck变换器的模糊规则,可以使控制器不依赖于Buck变换器系统精确的数学模型,从而克服传统PID控制导致的输出电压高超调、振荡等缺点;最后,设计了一套新型双闭环Buck变换器硬件系统,利用STM32单片机将模糊PID算法首次用于控制变化速度快、时滞小的被控量。实验结果表明：模糊PID控制策略不仅能够提高输出电压跟踪精度,还能有效抑制负载扰动及参数摄动,在阶跃启动下,系统输出电压超调量低于10%,调节时间低于3ms;与传统PID单环系统相比,其输出电压纹波、扰动下的最大动态压降和恢复时间均减小了一个数量级以上。

英文摘要：

A novel fuzzy PID control strategy of Buck converter system with double closed loop structures is proposed to solve problems that complex control objects are difficult of modeling and the artificial tuning experiences are lack,etc.The voltage and current loop structures are designed to improve both the anti-disturbance performance and the following performance of the system.A membership function that combines the Gauss function and the triangular function is selected to enhance the sensitivity of the controller to small error.Three sub inference devices are designed based on the PID parameters and the rule base of expert experiences,and the fuzzy rule of Buck converter is developed to make the designed controller independent of the precise mathematical model of Buck converter.So that it overcomes the high percent overshoot,oscillation and other shortcomings of the output voltage caused by traditional PID controls.A new Buck converter system of double closed loop circuits is designed,and the fuzzy PID control strategy is applied to control the quantities with rapid change and small time delay by making useof STM32.Experimental results show that the fuzzy PID control strategy not only improves the tracking accuracy of the output voltage,but also effectively suppresses the load disturbance and parameter perturbation,and that the overshoot of the output voltage is less than 10% and the adjustment time is less than 3ms under the step to start.A comparison with the traditional PID system show that the designed system＇s output voltage ripple,maximum dynamic voltage drop and recovery time under the disturbance are reduced by one order of magnitude.