基于随机森林的污水处理曝气系统研究与应用

Research and application of aeration system based on random forest for wastewater treatment

以某污水处理厂曝气系统为研究对象,采用描述性统计分析方法对其工艺运行参数变化规律进行了分析,以此为基础,采用随机森林算法建立了污水处理智能曝气系统预测模型,并基于所建模型构建了污水处理智能曝气控制系统。研究结果表明,该曝气系统预测模型所预测风机频率与实测结果相吻合,两者平均误差在2.9%以内,具有较高的准确性。现场应用结果表明,采用智能曝气控制系统的生物滤池溶解氧可稳定在目标值±0.2 mg/L范围内,与人工控制池组相比,在气水比基本相同的情况下,智能曝气系统控制池组末端出水氨氮相对较低。

Based on the aeration system of a wastewater treatment plant as the research object,the variation law of process operation parameters is analyzed by using descriptive statistical analysis method.On this basis,the prediction model of wastewater treatment aeration system is established by using random forest algorithm,and the intelligent aeration control system of wastewater treatment is constructed based on the model.The results show that the fan frequency predicted by the model is consistent with the measured results,and the average error between them is less than 2.9%.The field application results show that the dissolved oxygen of the biofilter used the intelligent aeration control system can be stable within the target value±0.2 mg/L.Compared with the manual control system,the ammonia nitrogen in the biofilter outlet using the intelligent aeration control system is relatively low under the condition of basically the same gas-water ratio.