基于数据挖掘技术的污泥厌氧消化模拟研究

发布时间：2018-06-14 14:16

本文选题：数学模型 + 数据挖掘　；参考：《北京工业大学学报》2016年12期

【摘要】：为了更好地模拟实际工程中污泥厌氧消化系统的产气效果,以北京某大型污泥厌氧消化工程为例,以大量的工程数据为基础,分别采用多元线性回归模型、神经网络模型、分类回归模型和邻近算法模型等数据挖掘技术,对系统的产沼气效率进行了模拟预测,其中邻近算法模型具有最好的拟合效果.对邻近算法模型进行进一步研究分析,通过交叉验证法近一步优化了模型k值的选取,从测试结果可以看出随着k值增加,训练集的拟合度先下降后趋于平稳,测试集的拟合度则相反.最终确定当k值取5时,模型预测值与实际值的相关度达0.862,优于系统默认参数下的拟合效果.试验证明:数据挖掘技术可以很好地应用于污泥厌氧消化工程的模拟计算,对于数学模拟在污水处理领域的应用具有一定指导意义.
[Abstract]:In order to better simulate the gas production effect of sludge anaerobic digestion system in practical engineering, taking a large sludge anaerobic digestion project in Beijing as an example, the multivariate linear regression model and neural network model were adopted based on a large amount of engineering data. Data mining techniques such as classification regression model and adjacent algorithm model are used to simulate and predict the biogas production efficiency of the system. The neighbor algorithm model has the best fitting effect. Through further research and analysis of the adjacent algorithm model, the selection of the model k value is optimized by the cross validation method. From the test results, it can be seen that with the increase of k value, the fitness of the training set decreases first and then tends to steady. The fitness of the test set is the opposite. Finally, it is determined that the correlation between the model prediction value and the actual value is 0.862 when k value is 5, which is better than the fitting effect under the default parameters of the system. The experiment shows that the data mining technology can be applied to the simulation calculation of sludge anaerobic digestion project and has certain guiding significance for the application of mathematical simulation in the field of sewage treatment.
【作者单位】：北京工业大学建筑工程学院;北京城市排水集团有限责任公司;
【基金】：国家水体污染控制与治理科技重大专项资助项目(2014ZX07201-001)
【分类号】：X703.1

【参考文献】