【作者】 赵麟菱；

【导师】 周立芳；

【作者基本信息】 浙江大学 ， 系统工程， 2007， 硕士

【摘要】 活性污泥法是当前最常用的一种污水处理方法，国际水协会(IWA)针对活性污泥法提出了ASM系列机理模型。ASM系列中的ASM2模型包含了活性污泥过程的去除COD，BOD和脱氮除磷的机理过程，比较完整的描述了活性污泥过程的污水处理机理，广泛应用于工业和市政污水处理。因此基于ASM2模型研究污水处理过程模型简化及优化控制策略具有重要的理论研究意义和实际应用价值。本文综述了活性污泥法模型的研究现状和活性污泥法应用于工业控制的现状，描述了污水处理的ASM2模型，并以一个典型的污水处理工艺A~2／O工艺处理过程作为背景，将ASM2模型应用于A~2／O过程中，建立起A~2／O过程的整体系统的机理模型；分别采用时标分解算法和聚类及PLS算法，对ASM2模型进行简化研究，由时标分解方法得到一个较为简化的非线性模型，由聚类和PLS算法对模型进行简化得到一组适用于模型预测控制的多线性模型；然后，以溶解氧浓度和污泥回流量作为操作变量，以出水的硝态氮和氨氮的浓度作为输出变量，将多线性模型作为预测模型，采用广义预测控制算法对该A~2／O工艺过程的出水水质进行设定点控制和区间控制；最后，结合大量国内外文献中的资料，得出了一种衡量典型污水处理过程的经济性能指标，并将该经济性能指标融入广义预测控制的目标函数中，得到一个双层的优化过程，上层对出水水质进行区间控制，得到的输出作为约束加入到下层的考虑经济性能指标的预测控制中，最后得出输入变量的序列。从仿真结果可以看出，考虑了经济性能指标的预测控制在经济成本上较为节约，并能达到水质标准，从而达到在满足出水水质要求的前提下尽量减小能耗。更多还原

【Abstract】 In the wastewater treatment processes, the activated sludge method is a mostly useful method. The activated sludge model series, which are proposed by International Water Association, are the mechanism models corresponding to the activated sludge process. In the series, Activated Sludge Model No.2(ASM2) describes the dynamic processes of COD, BOD, nitrogen and phourospher removal in the activated sludge process. Therefore, the ASM2 is abroadly used in the industrial and municipal wastewater treatment.Firstly, the status quo of the activated sludge model and the industrial application ofthe activated sludge method is summarized. Secondly, the Activated Sludge Model NO.2(ASM2) is described detailedly, and based on ASM2, a whole mechanism model of a typical A~2/O treatment process is established. Thirdly, the reduction of ASM2 is researched. In allusion to the multi-scale of ASM2, we apply the time-scale decomposition method to get a reduced nonlinear model—a group of differential algebra equations. As to nonlinearity of ASM2, we introduce the clustering and PLS algorithm to get multi linear models suitable for model predictive control. Fourthly, the generalized predictive control(GPC) algorithm is applied in the A~2/O process, In the predictive control, dissolved oxygen concentration and sludge refluence are used as manipulated variables, the ammonia and nitrate concentration in the effluent are used as output variables, and the multi models from the third section are used as the predictive model. Lastly, based on many correlative papers, an index measures the economic capacity of the A~2/O process is obtained. The index is then absorbed in the objective function of the predictive control, and the objective is changed into the energy cost saving when the effluent quality is up to par. The objective function is constituted of two levels: the upper level is about the effluent quality, and the appropriate output is then used as a constraint of the lower level which is about the economic capacity. The simulation result shows that the predictive control considering economic capacity helps to save the energy cost in the wastewater treatment process while the effluent quality according with standard.更多还原