大数据技术在高校污水MBR膜寿命预测的应用——以西安思源学院为例

The MBR Membrane Life Prediction in Universities’ Waste Water Treatment through Big Data Technology—Taking Xi’an Siyuan University as an Example

摘要: 采用大数据均值聚类分析方法，将西安思源学院再生水厂的A2/O-MBR系统3520天的运转数据进行处理。按照“日工业透水率”的定义，计算每个膜池的日工业透水率，剔除五类异常点，得到1#、2#、3#膜池的有效日工业透水率天数分别为2474天、2725天、和2652天。将每25个有效日工业透水率划分为一个计算单元，计算单元的算术平均值和标准偏差。将每个膜池的计算单元的算术平均值按次序排列，并回归得工业透水率衰减方程。该直线方程的截距表示开始时工业透水率。该直线方程的负斜率意味着工业透水率随着操作时间的延长而不断衰减。根据每个膜池的工业透水率衰减方程，可以确定1#、2#、3#膜池的工业透水率年衰减率分别是4.36%/年、4.10%/年、和4.54%/年。

Abstract: The 3520 days operation data of the A2/O-MBR system of Xi’an Siyuan University has been processed using big data mean clustering analysis. The daily industrial permeability of each membrane pool was calculated according to the definition. After eliminating the five kinds of anomaly points, the effective daily industrial water permeability days of membrane tank 1#, 2# and 3# were 2474 days, 2725 days and 2652 days, respectively. The industrial water permeability of every 25 effective days was divided into a calculation unit, and the arithmetic mean and standard deviation of the unit were calculated. The arithmetic mean of the computing units for each membrane pool was arranged in order and returned to the industrial permeability decay equation. The intercept of the linear equation indicates the industrial permeability at the beginning. The negative slope of the linear equation implies that the industrial permeability is declining with the operating time. Based on the industrial permeability decay equation of each membrane pool, it can be determined that the annual industrial permeability rates of the 1#, 2#, and 3# membrane pools are 4.36%/Y, 4.10%/Y, and 4.54%/Y, respectively.