改进粒子群算法应用于锅炉汽包水位控制  

Water-level Control System of Boiler Drum Based on Improved Particle Swarm Optimization

作　　者：魏晋军[1] 张玮[1] 刘角 马迪[1] 

机构地区：[1]太原理工大学过程装备与控制工程系,太原030024

出　　处：《太原科技大学学报》2015年第6期423-429,共7页Journal of Taiyuan University of Science and Technology

基　　金：国家青年基金(51106104)

摘　　要：为提高粒子群算法的收敛速度,根据粒子位置期望的数学分析,提出一种基于高斯分布的粒子群优化算法,该算法利用高斯分布随机数选取的特性,若选择合适的参数,不仅有利于算法收敛速度增快,而且增加了跳出局部最优的机制。应用于四个标准测试函数表明,当高斯分布选取μ=0.5,P=97%(σ2=0.23)时,性能较优秀。之后将改进的粒子群算法应用于锅炉汽包水位串级三冲量PID控制系统的参数整定当中,经过matlab中的simulink模块仿真,取得较好的效果。An improved particle swam optimization with Gaussian distribution from mathematical expectation posi- tion of particles is proposed to improve the convergence rate of Particle Swarm Optimization. The algorithm uses the properties of Gaussian random number selection. If Gaussian distribution selects the appropriate parameters, the al- gorithm will not only make the algorithm convergence faster, but also provide a mechanism to help algorithm jump out of local optimum. Through the results applied to the four standard test functions, Gaussian distribution choose parameters μ=0.5,P=97%（σ2=0.23） can help algorithm obtain better performance. Then the improved Particle Swarm Optimization was used on PID parameter tuning,in which PID will be used for the boiler drum level control system. Better results were achieved through the establishment of system model and the simulink in matlab.

关 键 词：粒子群优化算法 收敛速度 高斯分布 锅炉汽包水位 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构] TP273.5[自动化与计算机技术—计算机科学与技术]