| t分布受控遗传算法优化BP神经网络的PM2.5质量浓度预测 |
| |
| 引用本文： | 荆涛,李霖,于文柱,王玉娟,郑永杰,田景芝.t分布受控遗传算法优化BP神经网络的PM2.5质量浓度预测[J].中国环境监测,2015,31(4):100-105. |
| |
| 作者姓名： | 荆涛 李霖 于文柱 王玉娟 郑永杰 田景芝 |
| |
| 作者单位： | 1. 齐齐哈尔大学 化学与化学工程学院,黑龙江省 齐齐哈尔,161006
2. 齐齐哈尔环境监测站,黑龙江 齐齐哈尔,161000 |
| |
| 基金项目： | 齐齐哈尔市科学计划项目(SFZD-2013176) |
| |
| 摘 要： | 根据齐齐哈尔大学监测点2014年3—5月PM2?5质量浓度及其对应的每小时的气象因素、气体污染物浓度，建立基于t分布受控遗传算法的BP神经网络模型（ BPM?TCG），对PM2?5质量浓度进行模拟预测。并将其与BP神经网络模型、遗传算法优化BP神经网络模型（ BP?GA）进行对比分析。3种模型预测结果表明：BPM?TCG模型预测精度最高，泛化能力最好。 BPM?TCG模型对PM2?5质量浓度的准确预测为预防和控制PM2?5提供依据。
|
| 关 键 词： | t分布 受控衰减 遗传算法 BP神经网络 PM2?5 预测模型 |
| 收稿时间： | 2014/10/16 0:00:00 |
| 修稿时间： | 2014/12/28 0:00:00 |
Prediction of PM2.5 Mass Concentration Based on BP Neural Network Optimized by t-Distribution Controlled Genetic Algorithm |
| |
| JING Tao,LI Lin,YU Wen-zhu,WANG Yu-juan,ZHENG Yong-jie and TIAN Jing-zhi.Prediction of PM2.5 Mass Concentration Based on BP Neural Network Optimized by t-Distribution Controlled Genetic Algorithm[J].Environmental Monitoring in China,2015,31(4):100-105. |
| |
| Authors: | JING Tao LI Lin YU Wen-zhu WANG Yu-juan ZHENG Yong-jie and TIAN Jing-zhi |
| |
| Institution: | College of Chemistry and Chemical Engineering, Qiqihar University, Qiqihar 161006, China,College of Chemistry and Chemical Engineering, Qiqihar University, Qiqihar 161006, China,Qiqihar Environmental Monitoring Centre, Qiqihar 161000, China,Qiqihar Environmental Monitoring Centre, Qiqihar 161000, China,College of Chemistry and Chemical Engineering, Qiqihar University, Qiqihar 161006, China and College of Chemistry and Chemical Engineering, Qiqihar University, Qiqihar 161006, China |
| |
| Abstract: | According to the date of PM2.5 mass concentration and the corresponding hourly meteorological factors,gas pollutant concentration from March to May in 2014 at Qiqihar University monitoring site,BP neural network model optimized by t-distribution controlled genetic algorithm (BPM-TCG) was established. BPM-TCG was applied to simulate and predict PM2.5 mass concentration,and a comparative analysis was made between BPM-TCG and BP neural network model,BP neural network optimized by genetic algorithm (BP-GA). The experimental results showed that BPM-TCG possessed the highest precision and the best generalization ability. The accurate prediction of BPM-TCG model on PM2.5 mass concentration provides valuable reference for the prevention and control of PM2.5 pollution. |
| |
| Keywords: | t-distribution controlled attenuation genetic algorithm (GA) BP neural network PM2 5 prediction model |
|
| 点击此处可从《中国环境监测》浏览原始摘要信息 |
| 点击此处可从《中国环境监测》下载免费的PDF全文 |
|
