改进粒子群的模糊神经PID控制研究与应用  

Application of Fuzzy Neural PID Control Based on Improved PSO

作　　者：贺师超[1] 孔峰[1] 杨琳[1] 

机构地区：[1]广西工学院电子信息与控制工程系,柳州545006

出　　处：《自动化与仪表》2012年第4期35-39,共5页Automation & Instrumentation

基　　金：广西研究生教育创新计划基金项目(2011105940811M01)资助

摘　　要：设计一款电子负载用于对稳压电源进行检测。在硬件方面,对DSP控制器的选择、电子负载各芯片的选取、键盘及显示电路的设计进行了详细的介绍;在软件方面,针对模糊神经PID控制算法易陷入局部极值和迭代过久的缺点,加入了改进的粒子群算法,并且在神经网络在线调整时加入训练参数,对算法进行了改进,提高了算法的效率。将设计好的电子负载对稳压电源源效应进行检测,取得了较好的结果。This paper describes an electronic load which was used to detect power supply. On the hardware side,the choice of DSP,the chips select of electronic load, keyboard and display circuit designs were described in detail;On the software side,it was easy to fall into local minima and the shortcomings of iterations for the fuzzy neural PID control algorithm. Author uses an improved PSO,and neural network training coefficients to improve algorithm and enhance the efficiency. After finished the design of electronic load,it was used to detect voltage-stabilized power supply and it was proved satisfied.

关 键 词：电子负载 稳压电源 源效应 模糊神经PID 改进粒子群 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]