【作者】 袁桂丽；

【导师】 刘吉臻；

【作者基本信息】 华北电力大学（北京） ， 热能工程， 2010， 博士

【摘要】 本文设计了几种基于免疫反馈原理的控制器和基于免疫原理的自适应免疫遗传算法,并在电站控制系统进行了大量的应用仿真研究。第一：通过免疫反馈原理的研究,将免疫反馈控制器与模糊控制、PID控制有效的结合,设计模糊自调整免疫增量PID控制,并将其与传统方法整定的PID控制在球磨机负荷控制系统进行仿真对比研究,将模糊免疫PID控制与解决大迟延的Smith控制、内模控制进行仿真对比实验。仿真结果表明,模糊免疫PID控制器具有很好的快速性和抗干扰性。第二：在充分分析不完全微分和串级控制内回路作用基础上,设计了免疫不完全微分PID一免疫P串级控制,并将其应用到电站给水系统中。仿真结果表明,该控制策略下,即使参数在很大范围变化,给水系统仍具有很好的快速性和具有很好的克服给水扰动、蒸汽扰动的能力,大大减小了虚假水位的影响。第三：设计模糊免疫Smith控制器,利用Smith控制解决大迟延问题,利用模糊免疫控制解决快速性和抗干扰性、模型不匹配时系统的稳定性问题。并将其应用于给水系统,仿真结果表明,模糊免疫改进Smith控制具有比Smith控制更好的快速性和抗干扰性。第四：设计了免疫内模控制器。利用免疫控制实现在线整定内模控制器的滤波器参数,解决内模控制系统快速性和鲁棒性之间的矛盾。并将其用于球磨机负荷控制系统,仿真结果表明,免疫内模控制比内模控制具有更好的快速性和抗干扰性。第五：设计了基于相似性矢量距的自适应免疫遗传算法。给出了自适应免疫遗传算法各个模块的设计方法,同时将自适应免疫遗传算法同遗传算法收敛性函数验证做了仿真实验对比,仿真结果表明,自适应免疫遗传算法较遗传算法具有更好、更快的寻优能力。第六：将自适应免疫遗传算法用于电站主汽温控制系统PID参数的优化和电站机组的经济负荷优化分配。仿真结果表明,自适应免疫遗传算法具有比遗传算法更强的寻优能力,尤其它的全局、快速收敛性能,为电站系统在线实时优化运行的实施,提供了有利的保证。更多还原

【Abstract】 Artificial immune controllers are designed based on immune feedback. And Adaptive Immune Genetic Algorithm (AIGA) are given via immune principle, and applied in power plant control systems by simulation.First:Fuzzy immune self-tuning PID control is addressed by considering immune feedback control, fuzzy control and PID control. And it is proved to have good rapidity and anti-disturbance by using simulation in mill load control system.Second:Based on the full analysis of incomplete differential and the inner loop role of cascade control, immune incomplete differential PID-immune P cascade control is designed and applied to feed water system in power plant. Simulation results show that, the feed water system adopting the control strategy has good rapidity, and has ability to overcome the feed-water disturbance and the steam disturbance. And the false level greatly reduces.Third:Fuzzy immune-Smith control is addressed in this paper. Smith control can be used to solve large delay. Fuzzy immune control can solve the rapidity, anti-disturbance, and system stability when model does not match. Applied to feed water system, simulation results show that the improved fuzzy immune-Smith control has better rapidity and anti-disturbance than Smith control.Fourth:Immune internal model controller is designed. Combining immune control and Internal Model Control, filter parameters can be tuned online, and the contradiction between Internal Model Control and robustness is solved. This controller is applied to simulate for mill load control system. Simulation results show that immune internal model has better rapidity and anti-disturbance than Internal Model Control.Fifth:Based on density regulation of biological immune theory, diversity preservation strategy and immune memory function, Adaptive Immune Genetic Algorithm is proposed. Simulation results show that Adaptive Immune Genetic Algorithm is better than genetic algorithm on searching ability.Sixth:The Adaptive Immune Genetic Algorithm is used for PID parameters optimization for main steam temperature control system in power plant, and load dispatching optimization of power plant unit. Simulation results show that Adaptive Immune Genetic Algorithm improves the convergence rate, and maintains the antibodies diversity. Adaptive Immune Genetic Algorithm is stronger than Genetic Algorithm on optimization capability and convergence speed. That is helpful for on-line optimization in power plant.更多还原