Prediction of air quality for sustainable environment by artificial intelligent techniques

Air pollution is a growing problem arising from high density of vehicle traffic, global heating, new power generation plants, and expanding industrial activities. Air quality index (AQI) was developed and introduced as a key tool to provide information on local air quality of a region for public health concerns. Monitoring and forecasting air quality in the urban area are important due to health related issue. Air quality prediction is a complex problem, artificial intelligent techniques are successfully used in modelling the complex and non-linear prediction problems. In this study, an adaptive neurofuzzy logic method was developed to estimate the impact of certain factors on air quality and pollution level in Jeddah. Data on the concentrations of six environmental pollutants such as SO₂, CO, PM₁₀, O₃, NO_x and H₂S were employed to predict the Air quality index (AQI) in the atmosphere. Using Air quality standards of Saudi Arabia, AQI was calculated by considering the pollutants' concentration as independent variables. Different weighting factors were calculated for each pollutant according to their importance and concentration. The performance of ANFIS model was assessed by a number of checking data for different stations established by Saudi Presidency of meteorology and environment to observe the pollutants' in Jeddah. The outcomes of ANFIS model were evaluated by fuzzy quality charts and compared to the results obtained from US-EPA air quality standards. According to the results of present study, fuzzy rule based ANFIS model is a comprehensive tool for prediction and assessment of air quality and tends to produce accurate results.