【作者】 冯严冰；

【导师】 周莉；

【作者基本信息】 安徽理工大学 ， 控制理论与控制工程， 2008， 硕士

【摘要】 工业锅炉是一个多输入、多输出、多回路的强非线性控制对象。汽包水位是工业锅炉安全、稳定运行的重要指标,是自动控制的重要环节。经过多年的理论研究,目前已形成多种控制策略。但由于系统存在惯性滞后、非线性、时变等特点,采用传统的控制理论无法取得理想的控制品质。本文基于免疫系统的调节机理提出了一种免疫优化算法和一种新型免疫PID控制器,并把它们应用于工业锅炉汽包水位控制系统,以改善控制的品质。概括介绍了免疫学知识和目前主要的免疫学习算法,分析了遗传算法早熟收敛的机理。在此基础上,模拟生物体的抗原识别、免疫应答等实际免疫行为,提出一种新的人工免疫算法。从理论上证明了算法的全局收敛性,并用两个测试函数对算法的性能进行了测试,仿真结果表明,该算法可以有效改善未成熟收敛的缺陷,提高全局和局部搜索能力,证明了算法的有效性。探讨生物免疫系统的调节规律,并对免疫系统复杂精细的自我调节机制进行抽象和简化,本文根据免疫的反馈模型,通过转换形成了免疫控制器。但通过对二阶对象的仿真表明,该控制器在调节过程中存在静差。把免疫控制器与常规PID控制器结合提出一种并联型的免疫PID控制器,对二阶和高阶对象仿真结果表明,免疫PID控制器具有较小的波动和较短的过渡过程时间,性能要优于常规PID控制器和免疫控制器。最后介绍了工业锅炉的基本知识,对汽包水位的动态特性与控制方案给予了详细的分析与论述,在此基础上,针对670t/h流量的锅炉的汽包水位控制系统,分别进行了常规PID控制器、免疫控制器和参数免疫优化的免疫PID控制器的仿真分析,结果表明,参数免疫优化的免疫PID控制器的控制品质优于常规PID控制器和免疫控制器。本文将免疫优化理论与具体控制对象相结合,以MATLAB为仿真平台,详细研究了免疫优化算法及免疫PID控制器在锅炉汽包水位控制系统中的应用,为大滞后、非线性系统提供了一种解决方式,具有一定的理论意义与应用价值。更多还原

【Abstract】 Industrial boiler is a severe nonlinear control object with multi-input, multi-output and multi-loop. Drum water level of boiler is an important indicator for safe and reliable operation of industrial boiler. After many years of theory research, many control strategies have been formed. But because the system has the characters of inertial delay, nonlinearity, time-varying, it is difficult to construct its accurate mathematical model. Traditional control theory can’t get the good control performance. This dissertation put forward a immune optimization algorithm and a new-style immune PID controller based on adjustive mechanism of immune system,and they are apply to control system of drun water level of boiler to improve the control quality.This dissertation generalize the knowledge of immunology and the resent main immune algorithms,and analyse the mechanism of earliness convergence of genetic algorithm.Then simulating the antigen identify of organism,immune response and other actual immune behavior, a new artificial immune algorithm is put forward. This dissertation prove the gobal astringency of the algorithm from the theory, and two testing functions are used to test the capability of the algorithm, the result of simulation indicate that the algorithm can improve the earliness convergence and advance the global and local searching ability,indicating the validity of the algorithm.Discussing the adjustive rule of the immune system, abstract and predigest the complicated self-adjustive mechanism of the immune system.Based on the immune recuperative model, a new immune controller is formed. But the simulation of the second order objiect indicate this controller exist steady-state error during the adjustive process. So combining the simple PID controller and the immune controller, a parallel connection immune PID controller is put forward. The simulation of the second order objiect indicate this controller has smaller fluctuatation and short transient process time,so the performance of the immune PID controller is better than the immune controller and the simple PID controller.At last,the essential knowledge of industrial boiler is introduced, the dynamic characteristis and control precept of the drum water level of boiler is detailedly analysed and discussed.Based on this,aiming at the drum water level of boiler ,which flux is 670t/h,the simulation is repectively done based on the simple PID controller, the immune controller and the immune PID controller. The result of simulation indicate the control quality of the immune PID controller is better than the simple PID controller and the immune controller.Combing Artificial Immune Control theory and concrete control object, Adopting the MATLAB software as simulation platform, This dissertation studied the immune control algorithm and immune PID controller on drum water level of boiler, provided a way to solve the system with great delay and nonlinearity, so it has some theory value and application value.更多还原