污泥膨胀动力学特性分析与智能特征建模

项目名称： 污泥膨胀动力学特性分析与智能特征建模

项目编号： No.61203099

项目类型： 青年科学基金项目

立项/批准年度： 2013

项目学科： 自动化学科

项目作者： 韩红桂

作者单位： 北京工业大学

项目金额： 24万元

中文摘要： 污泥膨胀是污水处理过程中经常发生的一种现象，已成为制约活性污泥工艺发展的重大难题之一。由于受进水水质、环境状况、运行条件等因素的影响，污泥膨胀动力学特性非常复杂，建模十分困难。本课题在深入分析活性污泥沉降过程动力学特性的基础上，剖析污泥膨胀的致因机理，研究污泥沉降过程变量与污泥膨胀之间的相关性，获取污泥膨胀的特征变量；对于不能在线测量的特征变量，结合主元分析等方法从相关的可测参量集中找出主元参量，建立其软测量模型。同时研究自组织递归神经网络，获取能够反映污泥沉降比(SV)、污泥体积指数(SVI)等变量的自组织特征模型，并分析自组织特征模型的精确度和稳定性。最后形成具有自主知识产权的污泥膨胀智能特征模型，解决污泥膨胀识别与预测问题。研究工作对于污水处理厂的优化设计，污水处理过程的实时控制，都具有非常重要的支撑作用。因此，本课题的研究具有较高的理论研究价值，成果具有广阔的应用前景。

中文关键词： 污泥膨胀；特征建模；神经网络；软测量；识别与预测

英文摘要： Sludge bulking, frequently occurred in the operation of wastewater treatment processes (WWTPs), is one of the most serious problems that significantly restrain the development of activated sludge processes. Sludge bulking results from the influent water qualities, the environmental situations, the operating conditions and some other factors. Therefore, it is very difficult to analyze the dynamic characteristics and model the mechanism for sludge bulking. This project will investigate the dynamic characteristics of the activated sludge settling process, analyze the reasons of sludge bulking, study the correlation between settling process variables and sludge bulking, and obtain sludge bulking characteristic variables. For the immeasurable characteristic variables, the soft sensors will be proposed and modelled using the principal variables from the measurable parameters set based on the principal components analysis and other methods. Meanwhile, the self-organizing recurrent neural network will be studied. Then the self-organizing characteristic models will be obtained for the settling velocity (SV), sludge volume index (SVI) and the other parameters. In addition, the accuracy and stability of the self-organizing characteristic models will be discussed. Finally, the intelligent characteristic model for sludge bul

英文关键词： Sludge bulking；Characteristic model；Neural network；Soft measurement；Identification and prediction