循环流化床燃烧对象动态辨识与模糊自适应控制的研究

发布时间：2018-01-02 02:03

本文关键词：循环流化床燃烧对象动态辨识与模糊自适应控制的研究　出处：《东南大学》2016年硕士论文　论文类型：学位论文

【摘要】：循环流化床锅炉燃烧技术是近几十年发展起来的一项洁净煤燃烧技术。它燃烧效率高,燃料适应性广,污染排放低,和煤粉炉相比具有独特的优势。燃烧控制系统作为CFB锅炉安全、经济、稳定运行的保障,显得格外重要,是众多学者研究的重点。但是由于其非线性、参数时变、多变量耦合等特点,常规控制方法的效果难以令人满意。本文将具有高效搜索性能的蚁群算法和具有分布性、鲁棒自适应性的免疫系统机制引入热工过程控制领域,为解决复杂对象的辨识建模和动态自适应控制研究提供了一个崭新的思路。具体内容如下：(1)在分析基本蚁群算法工作原理与特点的基础上,引入免疫系统的抗原识别、抗体多样性、浓度调节等机制,从信息素初始分布、信息素调整机制、选择概率函数等方面对基本蚁群算法加以改进,增强算法的初始搜索效率与调整信息素的能力,提高了算法的自适应性与鲁棒性。通过对测试函数的求解和简单热工对象的验证,表明改进蚁群算法与PSO等其他智能算法相比具有更好的寻优特性和辨识能力。(2)在分析CFB锅炉燃烧系统的控制目标与特点的基础上,明确控制任务与对象,深入了解锅炉床温、床压的控制特点与难点。采取分散控制策略,只考虑回料阀开度和一次风量这两个主要影响因素,建立被控对象模型。通过对回料阀开度和一次风量的作用机理研究,确定床温床压传递函数模型结构。进而运用改进蚁群算法对其进行辨识。结果表明该模型能够很好地反映对象的变化趋势,基本满足工程应用需求。(3)引入模糊变论域思想,设计了自适应直接型模糊控制器,通过调整量化因子和比例因子使得模糊控制器的论域能够自适应改变。同时运用模糊规则调整函数,使模糊规则表能够根据上文得到的传递函数模型参数而自适应调整,克服了传统建立模糊规则表方法的缺陷。最后对燃烧控制系统进行了给定值扰动、内扰、外扰等情况下的仿真。结果表明,自适应模糊控制器比常规PID控制器具有更好的控制性能以及抗干扰的能力。
[Abstract]:The combustion technology of circulating fluidized bed boiler is a clean coal developed in recent years. Its combustion technology with high combustion efficiency, fuel adaptability, low emissions, and compared with the pulverized coal boiler has unique advantages. As CFB boiler combustion control system safety, stable operation of the economy, security is particularly important, is the focus of many scholars study. But because of the nonlinear, time-varying parameters, multivariable coupling, the effect of conventional control methods is not satisfactory. This paper is an efficient search and ant colony algorithm performance is distributed, robust and adaptive mechanism of the immune system of the introduction of thermal process control field, provides a new way for the Research of control dynamic modeling and adaptive solution of complex object. The specific contents are as follows: (1) based on analyzing the principle and characteristics of the basic ant colony algorithm on the introduction of immune system Antigen recognition system, antibody diversity, concentration regulation mechanism, from the initial distribution of pheromone, pheromone adjustment mechanism, selection probability function improved the basic ant colony algorithm, enhance the ability of the algorithm the initial search efficiency and adjust the pheromone, improves the adaptability and robustness of the algorithm. Through the verification to solve test functions and simple thermal objects, show that the improved ant colony algorithm and PSO algorithm are better than other intelligent optimization characteristics and identification ability. (2) based on the control objectives and characteristics analysis of CFB boiler combustion system, define the task and object, in-depth understanding of the boiler bed temperature, control characteristics and difficulties bed pressure. A decentralized control strategy, considering only the return valve opening and a volume of the two main factors, establish the model of the controlled object. Through the open degree and effect of primary air volume on the return valve Study on the mechanism, determine the bed temperature and pressure transfer function model structure. Then use the improved ant colony algorithm to identify. The results show that the model can well reflect the changing trend of the object, basically meet the engineering requirement. (3) the introduction of fuzzy variable universe theory, design the direct adaptive fuzzy controller, by adjusting the quantization the domain factor and proportional factor fuzzy controller can adaptively change. At the same time using fuzzy rules adjustment function, make the fuzzy rules according to the transfer function model parameter and adaptive adjustment of the above, overcome the defects of the traditional method of fuzzy rule table is established. At the end of the combustion control system of a given value disturbance, disturbance, simulation external disturbance conditions. The results show that the control performance of the adaptive fuzzy controller is better than conventional PID controller and anti-interference ability.

【学位授予单位】：东南大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TK229.66

【相似文献】