废水处理系统多元智能-机理模型耦合机制及动态模拟

Due to the increased concern about environmental issues, many experts focus their attention on the proper operation and control of wastewater treatment process (WWTP). In order to improve operating efficiency of wastewater treatment process and realize the wastewater treatment control, it is the base and key issue to build mathematical model of wastewater treatment process. However, operation of WWTP affected by a variety of physical, chemical, and biological factors, exists high nonlinearity, uncertainty and posterity, so establishing accurate mathematical model which describes wastewater treatment process is a difficulty and focus in the control field of wastewater treatment process..For the purpose of promoting the activated sludge model to be applied in control and optimization fields, singular perturbation will be proposed to decompose the mechanism model of wastewater treatment system and reduce the system dimension by analyzing dynamic characteristics of the wastewater treatment process, and then a simplified mechanism model presented as singular perturbation model will be derived based on nature of the mechanism model. Moreover the coupling mechanism between the simplified model and intelligent model will be discussed according to the performance of some intelligent algorithm (neural network, fuzzy logic, genetic algorithm, Markova chain, etc.), and a multiple intelligent- mechanism model will be developed to realize the dynamic simulation of the wastewater treatment process and clarify the internal relations between the dynamic parameters and degradation of pollutions. On this basis, combining embedded technology, a simulation program will be designed based on Matlab/Simulink and the dynamic simulation system of the multiple coupling model will be set up, which can reflect the dispersion and migration of pollutant visually, obtain the optimization and control strategy of the wastewater treatment process system. The research will provide theoretical guidance and technical support for impressive understanding the dynamic behavior and regularity of wastewater treatment system, stable running and process optimization of the wastewater treatment process system.

废水处理系统数学模型是提高废水处理系统运行效率和管理水平的重要工具，由于废水处理系统具有高度的多变量、非线性、时变性等特点，所以构建准确而实用的废水处理系统数学模型一直是废水处理控制技术领域的热点与难点。本项目通过分析废水处理系统的动态特性，运用奇异摄动理论对废水处理机理模型进行时标分解，建立简化机理模型使系统降维，并探讨其与神经网络、模糊逻辑、遗传算法、马尔可夫链等多种智能理论的耦合机制，构建机理与多种智能算法为一体的多元耦合动态模型，实现对废水处理系统的动态模拟，阐明动态变化参数与污染物降解过程的内在关系。在此基础上，结合嵌入式技术，在matlab/simulink平台编写模拟程序及搭建多元耦合模型动态仿真系统，以直观反映污染物随时间和空间的变化历程，优化和调控废水处理系统。本研究为深入理解废水处理系统的动态行为和规律、优化废水处理系统设计和运行方案等提供理论指导和技术支持。

近年来我国水污染问题尤为突出，治理任务相当繁重。为了更高效地治理废水污染，人们已逐步将先进的计算机技术和自动化技术应用于污水处理的过程控制中，构建各种高性能废水处理控制系统，以保证处理过程稳定、可靠、安全，提高出水水质，降低能耗、药耗和人工成本，从而实现污水处理的稳定、经济和良性运行。传统控制理论由于无法逼近复杂的非线性关系等自身的缺陷，控制效果也不尽人意，智能控制是控制理论发展的高级阶段，包括神经网络、模糊控制、专家控制等，可以根据废水水质参数的时变性，准确控制废水处理过程中的各项工艺参数，达到最好的处理效果，是当前污水处理工业控制领域一项备受关注的研究课题。.本研究通过分析废水处理系统的动态特性，运用奇异摄动理论对废水处理机理模型进行时标分解，并探讨了其与神经网络、模糊逻辑、遗传算法、马尔可夫链等多种智能理论的耦合机制，构建起机理与多种智能算法为一体的多元耦合动态模型，实现对废水处理系统的动态模拟。在此基础上，结合嵌入式技术，在matlab/simulink 平台编写模拟程序及搭建起多元耦合模型动态仿真系统，搭建起多元耦合智能优化控制系统，可直观反映污染物随时间和空间的变化历程，得到了调控和优化废水处理系统的策略。本研究为深入理解废水处理系统的动态行为和规律、优化废水处理系统设计和运行方案等提供理论指导和技术支持。