...
  • 主题
高级搜索  >
历史搜索 清空历史
首页> 外文期刊>Atmospheric research >The regional prediction model of PM10 concentrations for Turkey
【24h】

The regional prediction model of PM10 concentrations for Turkey

机译：土耳其PM10浓度的区域预测模型

获取原文
获取原文并翻译 | 示例
           
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

This study is aimed to predict a regional model for weekly PM10 concentrations measured air pollution monitoring stations in Turkey. There are seven geographical regions in Turkey and numerous monitoring stations at each region. Predicting a model conventionally for each monitoring station requires a lot of labor and time and it may lead to degradation in quality of prediction when the number of measurements obtained from any omonitoring station is small. Besides, prediction models obtained by this way only reflect the air pollutant behavior of a small area. This study uses Fuzzy C-Auto Regressive Model (FCARM) in order to find a prediction model to be reflected the regional behavior of weekly PM10 concentrations. The superiority of FCARM is to have the ability of considering simultaneously PM10 concentrations measured monitoring stations in the specified region. Besides, it also works even if the number of measurements obtained from the monitoring stations is different or small. In order to evaluate the performance of FCARM, FCARM is executed for all regions in Turkey and prediction results are compared to statistical Autoregressive (AR) Models predicted for each station separately. According to Mean Absolute Percentage Error (MAPE) criteria, it is observed that FCARM provides the better predictions with a less number of models. (C) 2016 Elsevier B.V. All rights reserved.
机译：这项研究旨在预测土耳其每周测量的PM10浓度的空气污染监测站的区域模型。土耳其有七个地理区域，每个区域都有许多监控站。常规地，为每个监视站预测模型需要大量的劳动和时间，并且当从任何监视站获得的测量数量较小时，这可能导致预测质量的下降。此外，通过这种方式获得的预测模型仅反映了小区域的空气污染物行为。本研究使用模糊C自回归模型（FCARM）来找到一个预测模型，以反映每周PM10浓度的区域行为。 FCARM的优势在于能够同时考虑指定区域中监测站的PM10浓度。此外，即使从监视站获得的测量数量不同或很小，它也可以工作。为了评估FCARM的性能，对土耳其的所有地区都执行了FCARM，并将预测结果与分别为每个站点预测的统计自回归（AR）模型进行了比较。根据平均绝对百分比误差（MAPE）标准，可以观察到FCARM用较少的模型提供了更好的预测。 （C）2016 Elsevier B.V.保留所有权利。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱：kefu@zhangqiaokeyan.com

京公网安备：11010802029741号 ICP备案号：京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有

  • 客服微信

  • 服务号