【作者】 刘超彬；

【导师】 乔俊飞；

【作者基本信息】 北京工业大学 ， 控制理论与控制工程， 2006， 硕士

【摘要】 活性污泥法污水处理过程是利用污泥中微生物的生命活动来清除污水中污染物质的一种有效方法。由于输入水质水量的剧烈变化,以及微生物生长的复杂性,使得该过程具有多变量、非线性、强耦合、大滞后和不确定性等特点,这导致了其过程控制自动化水平的相对落后。为保证处理过程运行良好和提高出水质量,开发和研究新型的控制策略,已经成为污水控制工程领域的重要课题。进行污水处理过程的智能控制研究,不仅可以强化和充实智能控制的理论和应用范围,而且有助于丰富和发展控制理论与方法,可以为非线性、大滞后的复杂系统控制问题提供新的理论方法。本文在深入分析现有研究成果的基础上,对活性污泥法污水处理过程的智能控制方法及仿真技术进行了研究。主要工作如下:1、对污水处理过程的自动控制现状进行了综述;2、在分析活性污泥污水处理系统的内在机理基础上,建立了反映溶解氧、活性污泥以及底物之间内在关系的数学模型;3、针对活性污泥系统的复杂性,结合目前智能控制研究的热点,设计了模糊神经网络控制器,并分别对溶解氧和污泥龄的控制进行数字仿真实验。①对序批式间歇活性污泥法以及完全混合式连续流活性污泥法中溶解氧浓度的控制设计了4层模糊神经网络控制器,详细分析了其结构及算法,并验证了该控制器的鲁棒性与自适应性。②对完全混合式活性污泥法进行了污泥龄的模糊神经网络控制,设计了5层模糊神经网络控制器,保持曝气池中污泥浓度在一定水平,使污泥具有良好的去污能力和沉淀性能。③与基于规则的模糊控制以及污水处理中的传统控制方式进行对比研究,结果表明具有学习能力的模糊神经网络控制应用于污水处理系统中可以获得更优的性能。本文的研究对我国污水处理过程的智能控制技术开发有一定借鉴意义。更多还原

【Abstract】 Activated sludge process is a kind of effective method, which consumes the pollutant in wastewater by microorganisms’metabolism. Due to the violent change of influent and the complexity of microorganisms’growth, the activated sludge processes have characteristics of muti-variables, strong coupling, high non-linearity, big lag and indetermination, which result in laggard automation of wastewater treatment process. To ensure that the treatment process run well and enhance effluent quality, it has become an important task in the field of wastewater treatment to research new control method. Researching the intelligent control strategy in wastewater treatment can not only enrich theory and application of intelligent control, but also is useful to the development of control theory, which can provide new control way for nonlinear and complicated systems.In the paper, on the basis of thorough analyzing past achievements, the author studied intelligent control technology of activated sludge wastewater treatment process. The main content of this paper is as follows:(1) Summarized the automatic control strategies in wastewater treatment.(2) Analyzed the inherent mechanism of activated sludge, established mathematics model, which reflects the relations among dissolved oxygen, microorganism and substrate.(3) Aiming at the complexity of activated sludge system, and combining the hotspot of intelligent control, we designed fuzzy neural network controllers, expatiated their structures and optimization algorithms, and respectively applied them to the control of dissolved oxygen concentration and sludge age.①Designed four layers fuzzy neural network to control the dissloved oxygen concentration in the sequencing batch reator and the complex mix bioreactor with continuous influent. The result of simulation experiments confirms the feasibility of control strategy and proves the robustness and self-adaptability of fuzzy neural network controller.②Designed five layers fuzzy neural network to control the sludge age in the complex mix bioreactor. The results of experiments indicated that the controller can keep the actived sluge concentration in reactor at a proper level and make sludge have good decontamination ability and settle capability.更多还原