改进的遗传算法在CFB锅炉热工过程建模中的应用  

Application of improved genetic algorithm to CFB boiler's thermal process modeling

作　　者：于希宁[1] 李亮[1] 范瑾[1] 

机构地区：[1]华北电力大学控制科学与工程学院,河北保定071003

出　　处：《华东电力》2007年第12期1-5,共5页East China Electric Power

基　　金：国家自然科学基金资助项目(50576022)

摘　　要：针对循环流化床(CFB)锅炉热工被控对象的特点及遗传算法存在的问题,提出一种用于热工过程建模的改进遗传算法,此算法引入模糊集理论,实现交叉概率和变异概率的模糊自整定,有效抑制了算法早熟,提高了算法的全局搜索能力。利用阶跃响应法获得现场特性曲线,基于模糊遗传算法得到典型负荷处的传递函数,将建模结果用于现场控制器的设计。对主汽温系统现场控制器进行内模控制整定,并进行了仿真研究和实际应用,结果表明该方案有较好的鲁棒性和抗干扰能力。Based on the characteristics of the thermal process of the circulating fluidized bed （CFB） boiler and problems in genetic algorithm, an improved genetic algorithm for thermal process modeling is proposed. In the algorithm, the fuzzy-set theory is introduced to realize fuzzy self-tuning of crossover probability and mutation probability, which effectively prevents premature convergence and improves its global search ability. The step response method was used to get the field characteristic curve, and the fuzzy genetic algorithm was applied to achieve the transfer function for typical loads. The algorithm was used for internal model control tuning of the main steam system field controller, and simulation and application prove that it has good robustness and strong anti-jamming capability.

关 键 词：CFB锅炉 模糊遗传算法 内模控制整定 仿真 

分 类 号：TK323[动力工程及工程热物理—热能工程]