【作者】 黄明智；

【导师】 万金泉；

【作者基本信息】 华南理工大学 ， 环境工程， 2010， 博士

【摘要】 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。本课题在较全面地分析了污水处理智能控制研究现状,以及模糊控制(FC)、神经网络(ANN)、遗传算法(GA)基本原理的基础上,以神经网络和模糊逻辑(FL)理论为核心提出了污水处理混合智能控制思想,将模糊控制与PID、神经网络及遗传算法相结合的混合智能控制方法引入本课题组研制开发的高效一体化废水处理工艺自动控制系统中,并用智能方法对废水中的典型有机物在此系统中的降解机理进行建模及对其运行状况进行评估,在污水处理混合智能控制跟建模研究方面进行了一些开拓性和探索性的研究工作,包括如下几个方面:(1)论文构建了基于Windows CE.NET嵌入式操作系统和MCGS组态软件的嵌入式废水处理自动控制系统,并在此基础上提出了Fuzzy-PID、GA-BP和改进模糊神经网络混合智能控制结构和设计方法。(2)针对污水处理厂水量、水质对控制系统稳定性的影响和目前水质在线检测仪器(仪表传感器)欠缺这一实际问题,广泛采用智能化建模技术,提出了利用将神经网络建模技术与模糊数学模型结合的水质多目标预测智能计算(软测量)模型与算法,并结合数据统计分析方法(PCA),估计出水水质考核指标,为解决水质测量滞后对污水处理控制系统带来的不平稳进行了创新性研究。(3)在本课题组自行开发的高效一体化混凝反应器的基础上,通过对废水处理混凝特征及模糊神经网络理论的分析,提出了基于聚类算法的混凝投药智能预测和控制模型。通过模糊C均值聚类方法从样本数据中总结出12条模糊规则,同时结合混合算法完成网络的结构辨识和参数辨识,最后将预测控制模型与MCGS组态挂接实现加药量的自动调节和出水COD跟踪期望COD为目标的反馈控制,能准确的控制加药量。(4)在对污水机理模型研究的基础上,针对机理建模难以应用于控制研究中的状况,提出活性污泥污水处理系统的一种变参数数学模型,可以反映出曝气量与溶解氧浓度的内在关系,并在对模型性能分析与综合的基础上,结合前馈与反馈控制方法,建立了溶解氧的两级模糊神经网络混合控制器,通过仿真试验,证明控制系统具有较好的快速性、稳定性,稳态性能以及鲁棒性,证实了控制方案的可行性与有效性。(5)分析确定了不同进水负荷在A/O系统的最佳回流比和反硝化反应对进水负荷利用效率,获得其内循环控制策略,并根据营养物质在系统中的变化规律,建立营养物浓度动态变化的模糊神经网络预测模型,可以跟踪营养物在A/O系统中的动态变化,最后以缺氧池末端的硝酸氮作为建模参数,进行模糊神经网络控制研究,实现反应器的脱氮处理控制。(6)通过测定邻苯二甲酸二丁酯(DnBP)在A2/O系统的三相有机物含量,确定不同性质有机组分迁移转化规律,研究其在厌氧、缺氧和好氧去除机理,建立起相应的动力学去除模型;同时在此基础上,通过各个参数之间的关系,运用神经网络和遗传算法理论,基于DnBP去除机理,建立了DnBP的GA-BP模型,实现DnBP在A2/O系统中的动态变化的预测,丰富了去除机理。本文对于智能控制技术在废水处理领域的深入研究和应用具有重要的参考价值。更多还原

【Abstract】 Recently wastewater pollution is very severe and the treating is still a heavy task. In order to treat wastewater effectively and realize recycling, it is important to develop computer operational decision support systems (computer technology and automation technology) that are able to play a similar role to the expert in wastewater treatment process control, in which it can ensure the operation safety, stable and reliable, and improve operational performance of wastewater treatment processes, so as to realize the wastewater treatment continuous, economic and benign operation. Previously, applications of control theory to wastewater treatment mainly focused on issues of nonlinearity, uncertainty and posterity, where difficulties in establishing accurate mathematical models and designing reliable controllers existed. Intelligent control is recently developed from conventional control theory. It consists of several control theories, such as fuzzy control (FC), artificial neural network control (ANNC), expert control, etc. The intelligent control of wastewater treatment system is a focus in wastewater treatment research field.Base on the all-around review and analysis of the progress of wastewater treatment study, a hybrid-intelligent control method was presented in line with neural network and fuzzy logic theory, in which FC, PID and GA were all taken into consideration. The intelligent-intelligent control was used in highly effective integration wastewater treatment control system; using intelligent methods, the degradation mechanism of typical organic matter was modeled and the operation status was assessed; and several valuable conclusions were reached.In this paper, according to the characteristics of wastewater treatment process, the automatic control system based on Windows CE.NET OS and MCGS software of wastewater was constructed , and the design method and the hybrid intelligent structure were proposed with the consideration of Fuzzy -PID, GA- BP and Fuzzy neural network control.The model construction techniques of ANN and fuzzy mathematics were combined to describe the wastewater treatment process, then a soft-computation model was built for water quality prediction and the influent water indexes were evaluated. Advanced neuro-fuzzy modeling, namely an adaptive network-based fuzzy inference system (ANFIS), was employed to develop models for the prediction of suspended solids (SS) and chemical oxygen demand (COD) removal of a full-scale wastewater treatment plant treating process wastewaters from a paper mill. In order to improve the network performance, fuzzy subtractive clustering was used to identify model’s architecture and optimize fuzzy rule, meanwhile principal component analysis (PCA) was applied to reduce the input variable dimensionality. Input variables were reduced from six to four for COD and SS models, by considering PCA results and linear correlation matrices among input and output variables. For this reason, the lag problem of measuring system is solved satisfactorily.A fuzzy neural network predictive control scheme for studying the coagulation process of wastewater treatment was given based on the characteristic of wastewater treatment and fuzzy neural network’s analysis. An adaptive fuzzy neural network was employed to model the nonlinear relationships between the removal rate of pollutants and the chemical dosages, in order to adapt the system to a variety of operating conditions and acquire a more ?exible learning ability. The system includes a fuzzy neural network emulator of the reaction process, a fuzzy neural network controller, and an optimization procedure based on a performance function that is used to identify desired control inputs. The gradient descent algorithm method was used to realize the optimization procedure. After fuzzy C-means clustering to identify models’architecture and a hybrid learning rule to identify parameters, the simulation indicated that the predictive model had good ability both in learning and generalization. Also, Fuzzy neural network algorithm with MCGS development package using VB program was developed,and then it was embedded into MCGS according to MCGS interface function criterion to achieve intelligent control system for wastewater treatment,so the control model could change the dosage according to different situationAccording to analyze the inherent mechanism of wastewater treatment process, mathematics model was established, which reflects the relations among dissolved oxygen, microorganism and substrate. And an integrated neural-fuzzy process controller was developed for studying the aeration of wastewater treatment. In order to improve the fuzzy neural network performance, the self-learning ability embedded in the fuzzy neural network model was emphasized for improving the rule extraction performance. The fuzzy neural network proved to be very effective in modeling the aeration performs better than artificial neural networks (ANN). For comparing between operation with and without the fuzzy neural controller, and comparing with operation with PID controller, a aeration unit in a Wastewater Treatment Process was picked up to support the derivation of a solid fuzzy control rule base. It is shown that, using the fuzzy neural controller, in terms of the cost effectiveness, it enables us to save almost 33% of the operation cost during the time period when the controller can be applied. Thus, the fuzzy neural network proved to be a robust and effective control tool, easy to integrate in a global monitoring system for cost managing.Analyzing and identifying the best reflux ratio and the denitrification efficiency of the influent loading under the different influent loading in A/O system, the inner loop control strategy was obtained; and on basis of the change law of nutrients in the system, fuzzy neural network models for forecasting dynamic change of nutrient concentration were designed, which can track the dynamic changes of the nutrients in the A/O system; taking nitrate nitrogen in the end of the anaerobic pond as modeling parameter, the fuzzy neural network control model to control nitrate nitrogen in the reactor is designed.The lab scale tests for the anaerobic/anoxic/oxic (AAO) process were carried out to investigate the removal and fate of di-n-butyl phthalate (DnBP) and optimum systematic operation parameters. Transfer and transformation of organic substances in the AAO system were especially concerned. The removal characteristics of phthalate esters were studied in anaerobic/anoxic/oxic (AAO) processes. A removal (biological degradation and sorption) model was formulated, and kinetic parameters were evaluated with batch experiments under anaerobic, anoxic, oxic conditions. Furthermore, the intelligent methods (neural networks and genetic algorithm) were used to model the relationships between phthalate esters and the water quality characteristic parameters, so as to realize to forecast the effluent quality of anaerobic/anoxic/oxic reactors of the AAO process.The research can provide guidance for the further study of intelligent control in the field of wastewater treatment and the popularization of wastewater treatment project with intelligent control.更多还原