摘要
针对超超临界机组的主汽温控制,提出了一种广义预测PID控制方法,该方法采用改进遗传算法对PID控制器参数进行在线优化.采用了一种基于广义预测控制性能指标的遗传算法寻优思路,建立了主、副回路PID参数优化模型;在选择、交叉和变异算子方面,初始种群设计方面和交叉、变异概率调整方面对遗传算法进行了改进.最后对广东潮州电厂某锅炉现场连续运行的历史数据进行了控制仿真,结果表明采用基于改进遗传算法的预测PID控制策略后,系统在动、静态特性和变负荷能力上均比常规串级系统效果更好.
According to the main steam temperature control of ultra-superoritical unit,an improved genetic algorithm optimizing predictive control performance based PID controller parameter is presented.Adopted a generalized predictive control based on the performance indexes of genetic algorithm optimization idea,a lord,deputy loop PID parameter optimization model is established.The genetic algorithm is improved from the selection,crossover and mutation,the design of initial population and crossover and mutation probability adjustment.Finally,a control simulation is conducted on the historical data of the continuous operation of Chaozhou Power Plant boiler;and the results show that the system of improved genetic algorithm based predictive PID control strategy performs better than those of conventional cascade system in dynamic and static characteristics and variable load ability.
出处
《武汉大学学报（工学版）》
CAS
CSCD
北大核心
2013年第3期386-392,共7页
Engineering Journal of Wuhan University
基金
国家自然科学基金资助项目(编号:61170024)
中国博士后科学基金项目(编号:2011M501234)
关键词
主汽温
串级PID控制
遗传算法
预测控制
main steam temperature
cascade PID control
genetic algorithm
predictive control
作者简介
方彦军（1957-），男，教授，博士生导师，研究方向为先进控制系统及应用，E-mail：yjfang@whu．edu．cn．
