Modelling and analysis of ozone concentration by artificial intelligent techniques for estimating air quality
Modelling and analysis of ozone concentration by artificial intelligent techniques for estimating air quality
复制标题
DOI：
10.1016/j.atmosenv.2016.11.030
复制
发表时间：
2017-02
影响因子：
5
通讯作者：
O. Taylan
中科院分区：
文献类型：
--
作者：
O. Taylan
臭氧浓度高是造成空气污染的一个重要原因，主要是因为臭氧在温室气体排放中的作用。臭氧是由低层大气中含有氮氧化物和挥发性有机化合物的光化学过程产生的。因此，监测和控制城市环境中的空气质量对于公共卫生来说是非常重要的。然而，空气质量预测是一个高度复杂和非线性的过程；通常需要考虑几个属性。人工智能(AI)技术可用于监测和评价臭氧浓度水平。本研究的目的是开发一种自适应神经模糊推理方法(ANFIS)来确定外围因素对空气质量和污染的影响，这是吉达市臭氧水平引起的一个问题。在大气条件下，臭氧浓度可作为预测空气质量的一个因子。使用沙特阿拉伯的空气质量标准，臭氧浓度水平是通过使用某些因素来模拟的，例如氮氧化物(NOx)、大气压、温度和相对湿度。因此，开发了一个ANFIS模型来观测臭氧浓度水平，并通过从沙特阿拉伯王国气象和环境保护总局建立的监测站获得的测试数据对模型的性能进行了评估。ANFIS模型的结果通过模糊质量图表重新评估，使用基于US-EPA空气质量标准的质量规范和控制限值。本研究的结果表明，ANFIS模型是估计和评估臭氧水平的综合方法，是产生更真实结果的可靠方法。
High ozone concentration is an important cause of air pollution mainly due to its role in the greenhouse gas emission. Ozone is produced by photochemical processes which contain nitrogen oxides and volatile organic compounds in the lower atmospheric level. Therefore, monitoring and controlling the quality of air in the urban environment is very important due to the public health care. However, air quality prediction is a highly complex and non-linear process; usually several attributes have to be considered. Artificial intelligent (AI) techniques can be employed to monitor and evaluate the ozone concentration level.The aim of this study is to develop an Adaptive Neuro-Fuzzy inference approach (ANFIS) to determine the influence of peripheral factors on air quality and pollution which is an arising problem due to ozone level in Jeddah city. The concentration of ozone level was considered as a factor to predict the Air Quality (AQ) under the atmospheric conditions. Using Air Quality Standards of Saudi Arabia, ozone concentration level was modelled by employing certain factors such as; nitrogen oxide (NOx), atmospheric pressure, temperature, and relative humidity. Hence, an ANFIS model was developed to observe the ozone concentration level and the model performance was assessed by testing data obtained from the monitoring stations established by the General Authority of Meteorology and Environment Protection of Kingdom of Saudi Arabia. The outcomes of ANFIS model were re-assessed by fuzzy quality charts using quality specification and control limits based on US-EPA air quality standards. The results of present study show that the ANFIS model is a comprehensive approach for the estimation and assessment of ozone level and is a reliable approach to produce more genuine outcomes.