污水处理过程中的絮凝状态识别及在自动控制中的应用  

作　　者：王雪 严良文[1] 

机构地区：[1]上海大学机电工程与自动化学院,上海200444

出　　处：《工业控制计算机》2025年第2期117-118,121,共3页Industrial Control Computer

摘　　要：针对工业污水处理中投药量无法根据污水水质进行动态调整的问题,开发了一套基于污水絮凝图像采集、特征提取与处理的智能加药控制系统,旨在实时监测污水絮凝状态并动态调整药剂投加量,从而达到节省药剂量的目的。该系统通过pH传感器控制污水的酸碱度在8.3左右,通过工业相机拍摄污水中的絮状物图像、采用数学形态方法,提取其絮体个数、平均面积、絮状物数量、分形维数等特征值,并通过主成分分析确定关键控制参数,确定以平均面积作为目标值控制PAC的投加量,实现投药闭环控制。为了验证控制方法的有效性,搭建了实验系统,并设计了三种工况下的污水实验。实验结果表明,该系统能有效提取絮体特征参数,实现污水絮凝的智能加药控制。In response to the issue of the inability to dynamically adjust the dosage of chemicals in industrial wastewater treatment based on wastewater quality,this paper develops an intelligent dosing control system based on the acquisition,feature extraction,and processing of wastewater floc images.The aim is to monitor the real-time floc status in wastewater and dynamically adjust the dosage of chemicals to achieve the goal of saving chemical doses.The system controls the acidity and alkalinity of wastewater to around 8.3 using a pH sensor.It captures images of flocs in wastewater using an industrial camera and employs mathematical morphology methods to extract features such as floc count,average area,floc quantity,fractal dimension,and texture features.Key control parameters are determined through principal component analysis.The dosage of polyaluminum chloride(PAC)is controlled targeting the average area,achieving closed-loop dosing control.An experimental system was constructed to validate the effectiveness of the control method,and wastewater experiments were designed for three operating conditions.

关 键 词：污水处理 絮体图像 特征提取 加药控制 

分 类 号：X703[环境科学与工程—环境工程]