华南理工大学学报(自然科学版) ›› 2009, Vol. 37 ›› Issue (5): 135-138.

• 计算机科学与技术 • 上一篇    下一篇

基于Elman神经网络的污染源数据预测

张齐1  许志坚1  赵坤荣2   

  1. 1. 华南理工大学 计算机科学与工程学院, 广东 广州 510006; 2. 环境保护部 华南环境科学研究所, 广东 广州 510655
  • 收稿日期:2008-05-21 修回日期:2008-07-18 出版日期:2009-05-25 发布日期:2009-05-25
  • 通信作者: 张齐(1963-),男,副教授,主要从事智能控制、信息处理、监控软件研究. E-mail:csqzhang@scut.edu.cn
  • 作者简介:张齐(1963-),男,副教授,主要从事智能控制、信息处理、监控软件研究.
  • 基金资助:

    广东省科技计划项目(200713030100001);广州市科技攻关计划项目(200523190191)

Prediction of Data from Pollution Sources Based on Elman Neural Network

Zhang Qi Xu Zhi-jian Zhao Kun-rong2   

  1. 1. School of Computer Science and Engineering, South China University of Technology, Guangzhou 510006, Guangdong, China; 2. South China Institute of Environment Science, Ministry of Environmental Protection, Guangzhou 510655, Guangdong, China
  • Received:2008-05-21 Revised:2008-07-18 Online:2009-05-25 Published:2009-05-25
  • Contact: 张齐(1963-),男,副教授,主要从事智能控制、信息处理、监控软件研究. E-mail:csqzhang@scut.edu.cn
  • About author:张齐(1963-),男,副教授,主要从事智能控制、信息处理、监控软件研究.
  • Supported by:

    广东省科技计划项目(200713030100001);广州市科技攻关计划项目(200523190191)

摘要: 为了给环境保护决策提供有价值的预测数据,提出利用Elman神经网络建立污染源数据预测模型的方法,以大气中的主要污染物SO2为例,用预测模型表征SO2的浓度和气温、相对湿度、风速、时间等影响因子及其历史数据之间的复杂关系.使用训练后的模型对数据进行模拟仿真,结果表明所建立模型的计算输出值与实际样本数据有着较好的一致性.模型预测效果优于基于BP神经网络的预测模型.

关键词: 污染源, 预测, Elman 神经网络

Abstract:

In order to provide valuable prediction data for the decision making of environmental protection, a model to predict the data of pollution sources is proposed based on Elman neural network. By using the proposed model and by taking SO2--the main pollutant in the atmosphere--as an object, the complex relationships between SO2 concentration and the factors, including the air temperature, the relative humidity, the wind speed, the time and the historical date, are investigated. After training, the proposed model is used to perform a simulation. The results demonstrate that the outputs calculated with the proposed model accord well with the sample outputs, and that the proposed model is more effective than the prediction model based on BP neural network.

Key words: pollution source, prediction, Elman, neural network