基于单输入模糊PID控制算法的Buck型DC-DC变换器设计
发布时间：2018-07-24 19:09
【摘要】：DC-DC变换器的动态性能对于电子设备的安全可靠运行至关重要,在大多数功率变换器应用中,通常在其额定工作点附近采用局部线性化方法建立小信号数学模型,并基于此模型设计线性控制器。但是Buck型DC-DC变换器具有强非线性,并且其传递函数含有与负载相关的极点,这给传统线性控制策略在变换器中的应用带来了挑战。本文设计一种基于单输入模糊PID (single-input fuzzy PID, SIF-PID)控制算法的快速响应低超调Buck型DC-DC变换器,控制算法以较小的硬件资源占有率使变换器获得优良的动态性能。首先,研究变换器系统主拓扑和控制回路各功能模块的工作原理,建立补偿前系统的传递函数；然后采用频域法对PID控制器进行参数整定,并设计模糊逻辑输入输出变量的隶属函数；其次,通过分析变换器系统响应曲线及其在模糊规则表中的映射,建立具有快速响应和低超调特性的二维Toeplitz型模糊规则表,得到双输入模糊PID控制器；最后,采用符号距离法对双输入模糊PID控制器进行降维,得到单输入模糊PID控制器以降低算法的硬件资源占有率,同时运用遗传算法进行参数寻优,保证降维得到的单输入模糊PID控制器具有和双输入模糊PID控制器等价的控制性能。本文基于FPGA搭建数字控制Buck型DC-DC变换器验证平台,对系统在不同条件下进行测试。在启动过程中,基于单输入和双输入模糊PID算法的变换器建立时间分别为1821μs和178μs,二者近似相等,而基于常规PID算法的变换器建立时间为252μs；当负载电流在0.8A和1.0A之间跳变时,基于单输入和双输入模糊PID算法的变换器同样具有近似相等的恢复时间与过冲,且小于采用PID算法的变换器恢复时间与过冲：当输入电压从2.0V到5.0V变化以及负载电流从0A到1.2A变化时,基于单输入模糊PID算法的变换器输出电压能够以小于1%的偏差稳定在1.8V,其电压纹波为26mV,满足变换器的设计规格和稳态性能要求。
[Abstract]:The dynamic performance of DC-DC converter is very important for the safe and reliable operation of electronic equipment. In most power converter applications, the local linearization method is usually used to establish a small signal mathematical model near the rated operating point. The linear controller is designed based on this model. However, Buck type DC-DC converter has strong nonlinearity and its transfer function contains load related poles, which brings challenges to the application of traditional linear control strategy in converters. In this paper, a fast response low overshoot Buck DC-DC converter based on single input fuzzy PID (SIF-PID) control algorithm is designed. First of all, the principle of the main topology and the function modules of the control loop of the converter system is studied, and the transfer function of the system before compensation is established, and then the parameters of the PID controller are adjusted by the frequency-domain method. The membership function of fuzzy logic input and output variables is designed. Secondly, by analyzing the response curve of converter system and its mapping in fuzzy rule table, a two-dimensional Toeplitz fuzzy rule table with fast response and low overshoot is established. The double input fuzzy PID controller is obtained. Finally, the dimension of the dual input fuzzy PID controller is reduced by the symbolic distance method, and the single input fuzzy PID controller is obtained to reduce the hardware resource share of the algorithm. At the same time, the genetic algorithm is used to optimize the parameters. It is guaranteed that the single input fuzzy PID controller obtained by reducing the dimension has the same control performance as the dual input fuzzy PID controller. In this paper, the verification platform of digitally controlled Buck type DC-DC converter is built based on FPGA, and the system is tested under different conditions. In the start-up process, the establishment time of the converter based on single input fuzzy PID algorithm and double input fuzzy PID algorithm is 1821 渭 s and 178 渭 s, respectively, which are approximately the same, while those based on conventional PID algorithm are 252 渭 s, and when the load current is between 0.8 A and 1.0 A, the time of setting up the converter is 1821 渭 s. The converter based on single input fuzzy PID algorithm and double input fuzzy PID algorithm also has approximately equal recovery time and overshoot. And the recovery time and overshoot of the converter using PID algorithm are smaller: when the input voltage changes from 2.0V to 5.0V and the load current changes from 0A to 1.2A, The output voltage of the converter based on single input fuzzy PID algorithm can be stabilized at 1.8 V with a deviation of less than 1%, and its voltage ripple is 26 MV, which meets the design specifications and steady state performance requirements of the converter.
【学位授予单位】：东南大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TM46
本文编号：2142354
[Abstract]:The dynamic performance of DC-DC converter is very important for the safe and reliable operation of electronic equipment. In most power converter applications, the local linearization method is usually used to establish a small signal mathematical model near the rated operating point. The linear controller is designed based on this model. However, Buck type DC-DC converter has strong nonlinearity and its transfer function contains load related poles, which brings challenges to the application of traditional linear control strategy in converters. In this paper, a fast response low overshoot Buck DC-DC converter based on single input fuzzy PID (SIF-PID) control algorithm is designed. First of all, the principle of the main topology and the function modules of the control loop of the converter system is studied, and the transfer function of the system before compensation is established, and then the parameters of the PID controller are adjusted by the frequency-domain method. The membership function of fuzzy logic input and output variables is designed. Secondly, by analyzing the response curve of converter system and its mapping in fuzzy rule table, a two-dimensional Toeplitz fuzzy rule table with fast response and low overshoot is established. The double input fuzzy PID controller is obtained. Finally, the dimension of the dual input fuzzy PID controller is reduced by the symbolic distance method, and the single input fuzzy PID controller is obtained to reduce the hardware resource share of the algorithm. At the same time, the genetic algorithm is used to optimize the parameters. It is guaranteed that the single input fuzzy PID controller obtained by reducing the dimension has the same control performance as the dual input fuzzy PID controller. In this paper, the verification platform of digitally controlled Buck type DC-DC converter is built based on FPGA, and the system is tested under different conditions. In the start-up process, the establishment time of the converter based on single input fuzzy PID algorithm and double input fuzzy PID algorithm is 1821 渭 s and 178 渭 s, respectively, which are approximately the same, while those based on conventional PID algorithm are 252 渭 s, and when the load current is between 0.8 A and 1.0 A, the time of setting up the converter is 1821 渭 s. The converter based on single input fuzzy PID algorithm and double input fuzzy PID algorithm also has approximately equal recovery time and overshoot. And the recovery time and overshoot of the converter using PID algorithm are smaller: when the input voltage changes from 2.0V to 5.0V and the load current changes from 0A to 1.2A, The output voltage of the converter based on single input fuzzy PID algorithm can be stabilized at 1.8 V with a deviation of less than 1%, and its voltage ripple is 26 MV, which meets the design specifications and steady state performance requirements of the converter.
【学位授予单位】：东南大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TM46
【参考文献】
相关期刊论文 前2条
1 许智榜;石晓瑛;;基于GUI的Buck变换器仿真研究[J];华东交通大学学报;2012年04期
2 常昌远;陈瑶;黄金峰;王青;;Buck型DC-DC变换器中单输入模糊控制器的设计与实现[J];东南大学学报(自然科学版);2012年02期
相关硕士学位论文 前1条
1 王伟威;适用于数字控制直流—直流变换器中的数字脉宽调制器的研究与实现[D];复旦大学;2012年
，本文编号：2142354
本文链接：https://www.wllwen.com/kejilunwen/dianlidianqilunwen/2142354.html
