|
|||||
|
|
| 基于BP神经网络的城镇污水厂活性炭自动投加系统研究 | |
| 引用本文： | 方荣业,史宇滨,蒋婷,李威,史惠祥.基于BP神经网络的城镇污水厂活性炭自动投加系统研究[J].浙江大学学报(理学版),2018,45(4):468-475. |
| 作者姓名： | 方荣业 史宇滨 蒋婷 李威 史惠祥 |
| 作者单位： | 1. 浙江大学 环境与资源学院, 浙江 杭州 310058; 2. 华东勘测设计研究院有限公司, 浙江 杭州 310014 |
| 基金项目： | 环保部水体污染控制与治理科技重大专项（2017ZX07206）. |
| 摘 要： | 近年来，活性炭吸附技术逐渐成为深度处理的主流技术，但粉末活性炭投加系统仍处于人工控制阶段，现场需要技术人员依靠经验确定活性炭投加量，易造成出水水质不稳定、活性炭药耗大等问题.以浙江省嘉善县某城镇污水处理厂的深度处理工艺为研究背景，以粉末活性炭投加系统为研究对象，针对活性炭投加控制系统滞后、非线性、复杂等问题，建立了BP神经网络前馈预测-PID反馈控制的自动投加控制系统.实践证明，该系统具有较强的自适应能力和较高的控制精度，出水COD达标率较人工控制提高了8.88%，活性炭日均消耗量削减了16.61%，取得了较好的经济效益. |
| 关 键 词： | 深度处理 粉末活性炭投加系统 BP神经网络 投药复合控制 |
| 收稿时间： | 2017-06-14 |
A study on the activated carbon intelligent dosing system for urban sewage treatment plants based on BP neural network. |
|
| FANG Rongye,SHI Yubin,JIANG Ting,LI Wei,SHI Huixiang.A study on the activated carbon intelligent dosing system for urban sewage treatment plants based on BP neural network.[J].Journal of Zhejiang University(Sciences Edition),2018,45(4):468-475. | |
| Authors: | FANG Rongye SHI Yubin JIANG Ting LI Wei SHI Huixiang |
| Affiliation: | 1. College of Environmental & Resource Sciences, Zhejiang University, Hangzhou 310058, China; 2. Huadong Engineering Corporation Limited, Hangzhou 310014, China |
| Abstract: | In recent years, adsorption of contaminants by activated carbon has become a predominant technology for intensive treatment of wastewater. However, the operation system is yet controlled manually. And the dose of activated carbon is determined by experience, resulting in instability of effluent COD value and excessive use of activated carbon. In order to solve the problems of long time lag, nonlinearity and complexity of the dosing system of activated carbon, based on BP network prediction plus PID feedback control, we developed an automatic dosing system of activated carbon as an intensive treatment technology of a municipal wastewater treatment factory in Jiashan county. The composite control system exhibits good self-adaptability and great control precision. The qualification percentage is enhanced by 8.88% than manual control, and consumption of activated carbon is decreased by 16.61% per day. |
| Keywords: | advanced treatment powder activated carbon dosing system BP neural network compound dosing control |
| 本文献已被 CNKI 等数据库收录！ | |
| 点击此处可从《浙江大学学报(理学版)》浏览原始摘要信息 | |
| 点击此处可从《浙江大学学报(理学版)》下载全文 | |
