Optimizing BP Networks by Means of Genetic Algorithms in Air Pollution Prediction
-
摘要: 大气污染预报可以对大气污染提出警示,保护人体健康和生活环境.为了对北京市PM10的质量浓度进行预报,建立了用于大气污染预报的遗传神经网络模型,该模型运用遗传算法优化神经网络的权值和阈值,有效提高了网络的收敛性和预报准确率.用改进后的神经网络对北京市PM10的质量浓度进行了模拟,并将模型模拟结果与美国第3代空气质量模型Models-3(CMAQ)的数值模拟结果进行了比较.实验结果表明:遗传神经网络模型和数值模型的模拟结果的平均相对误差分别为0.21和0.26,用于空气污染物质量浓度短期预报时,神经网络模型的预测精度与数值模型的预测精度相当.对于没有条件开展空气污染数值预报的城市或地区,神经网络是一种有效的替代方法.Abstract: Air pollution forecasting provides early warning before air pollution issue occurs,thus protects human health and living environment.A neural network model optimized by genetic algorithm was developed in order to predict PM10 concentrations in Beijing.The genetic algorithm was used to optimize the initial weights and threshold of the BP neural network in simulation.Astringency of network and accuracy of prediction were effectively improved.The improved network and Models-3 Community Multi-scale Air Quality(CMAQ) modeling system were both applied in the prediction of short-term PM10 concentration in autumn 2002 in Beijing.Resultsshowed good prediction capability of both models,and the mean relative errors were separately 0.21 and 0.26.When applied in short-term air pollution forecasting,neural network is of similar prediction accuracy compared with CMAQ.It is an effective alternate method for air pollution forecasting in areas where mathematical model on air pollution can't be widely applied.
-
Keywords:
- PM10 /
- neural network /
- genetic algorithm /
- CMAQ
-
-
[1] GRIVAS G,CHALOULAKOU A.Artificial neural network models for prediction of PM10hourly concentrations,in thegreater area of Athens,Greece[J].Atmospheric Environment,2006,40:1216-1229.
[2] JEF H,CLEMENS M,GERWIN D,et al.A neural network forecast for dailyaverage PM10concentrations in Belgium[J].Atmospheric Environment,2005,39:3279-3289.
[3] 黄牧涛,黄科.混合遗传神经网络算法研究及其在径流预报中的应用[J].长江科学院院报,2007,24(4):31-33.HUANG Mu-tao,HUANG Ke-lang.Daily runoff forecast system based on combined genetic algorithm and ANN[J].Journalof Yangtze River Scientific Research Institute,2007,24(4):31-33.(in Chinese) [4] 金龙,吴建生,林开平,等.基于遗传算法的神经网络短期气候预测模型[J].高原气象,2005,24(6):981-987.JIN Long,WU Jian-sheng,LIN Kai-ping,et al.Short term climate prediction model of neural network based on geneticalgorithms[J].Plateau Meteorology,2005,24(6):981-987.(in Chinese) [5] ARCHONTOULA C,MICHAELA S,NIKOLAS S.Comparative assessment of neural networks and regression models forforecasting summertime ozone in Athens[J].The Science of the Total Environment,2003,313:1-13.
[6] 白晓平,张启明,方栋,等.人工神经网络在苏州空气污染预报中的应用[J].科技导报,2007,25(3):45-49.BAI Xiao-ping,ZHANG Qi-ming,FANG Dong,et al.Application of artificial neural network to air pollution prediction inSuzhou City[J].Science&Technology Review,2007,25(3):45-49.(in Chinese) [7] CHEN D S,CHENG S Y,GUO X R,et al.An integrated MM5-CMAQ modeling approach for assessing trans-boundaryPM10contribution to the host city of 2008 Olympic summer games-Beijing,China[J].Atmospheric Environment,2007,41:1237-1250.
[8] CHEN D S,CHENG S Y,LI J B,et al.Application of LIDAR technique and MM5-CMAQ modeling approach for theassessment of winter PM10air pollution:a case study in Beijing,China[J].Water,Air and Soil Pollution,2007,181:409-427.
[9] CHENG S Y,CHEN D S,LI J B,et al.An ARPS-CMAQ modeling approach for assessing the atmospheric assimilativecapacity of the Beijing metropolitan region[J].Water,Air and Soil Pollution,2007,181:211-224.
[10] CHEN D S,CHENG S Y,GUO X R,et al.An integrated ARPS-MM5-CMAQ modeling approach for predicting PM10concentration in the metropolitan region of Beijing in winter[J].Environmental Informatics Archives,2005,3:439-448.
[11] 房小怡,蒋维楣,吴涧,等.城市空气质量数值预报模式系统及其应用[J].环境科学学报,2004,24(1):111-115.FANG Xiao-yi,JIANG Wei-mei,WU Jian,et al.Study on the development of numerical model system to predict urban airquality[J].Acta Scientiae Circumstantiae,2004,24(1):111-115.(in Chinese) -
期刊类型引用(2)
1. 计志威，刘安东，付明磊，孙虎，张文安，金锦峰，Ryhor Prakapovich，Uladzislau Sychou. 基于运动约束的无监督单目视觉里程计. 传感技术学报. 2024(09): 1545-1554 . 
百度学术
2. 杜慧丽，黄劲松. 利用掩膜和单应矩阵提高LK光流追踪效果. 导航定位学报. 2022(05): 25-31 . 百度学术
其他类型引用(1)
计量
- 文章访问数: 21
- HTML全文浏览量: 0
- PDF下载量: 6
- 被引次数: 3
下载:
