|
|||||
|
|
| PM_(10)污染及BP神经网络气象预测——以西安市为例 | |
| 引用本文： | 张强骅,姚锐,邓伟妮,王贵荣. PM_(10)污染及BP神经网络气象预测——以西安市为例[J]. 西安科技大学学报, 2010, 30(1) |
| 作者姓名： | 张强骅 姚锐 邓伟妮 王贵荣 |
| 作者单位： | 1. 青海岩土工程勘察咨询公司,青海,西宁,810001 2. 西安科技大学,地质与环境学院,陕西,西安,710054 3. 山东电力工程咨询院,山东,济南,250014 |
| 基金项目： | 陕西省自然科学基金项目 |
| 摘 要： | 在研究近几年西安市PM10污染的现状的基础上,初步选取8类20个气象因子,再采用主成分分析法进行精简,得到11个与PM10相关的主要因子,在此基础上,采用人工神经网络模型对西安市PM10污染状况进行预测,确定了网络模型结构。预测结果表明:预测值与实际值的相关系数达到0.801,在265个测试样本中,预测结果与实际完全吻合的为212d,占80%;相差不超过一级的天数为262d,占98.87%,与实际情况基本一致。 |
| 关 键 词： | 预测 BP神经网络 |
PM_(10) pollution and its BP neural network forecast model |
|
| ZHANG Qiang-hua,YAO Rui,DENG Wei-ni,WANG Gui-rong. PM_(10) pollution and its BP neural network forecast model[J]. JOurnal of XI’an University of Science and Technology, 2010, 30(1) | |
| Authors: | ZHANG Qiang-hua YAO Rui DENG Wei-ni WANG Gui-rong |
| Abstract: | The PM10 pollution of Xi'an city in recent years was introduced.Eight kinds,and twenty factors were selected primarily.These factors were chosen by principal component analysis and eleven main factors were obtained.The BP neural network model was built by means of software MATLAB and was used to forecast Xi'an PM10 pollution.The results indicated that the correlation coefficient of forecast value and actual value was 0.801,and of the 265 testing samples,the number of days when forecast value completely conf... |
| Keywords: | PM10 MATLAB |
| 本文献已被 CNKI 万方数据 等数据库收录！ | |