Vehicular pollution prediction using HWTO‐ANFIS model in urban areas of Hyderabad City

Abstract

AbstractAir pollution has become a life‐threatening factor around the world in recent decades as a result of vehicles, industrialization, human activities, and urbanization. The existing vehicular air pollution prediction studies met several shortcomings in terms of poor prediction accuracy, higher error rates, huge execution time and etc. To tackle these challenges, this study proposed a Hybrid Whale Tabu Optimization‐based Adaptive Neuro‐Fuzzy Inference System (HWTO‐ANFIS) predict the effects of vehicular pollution in urban areas. The major objective of this paper is to minimize the forecasting error and get optimal results during pollution prediction. The average, minimum and maximum noise levels of vehicles are effectively predicted using the proposed HWTO‐ANFIS model. The proposed model efficiency is evaluated using various performance measures such as Normalized Mean Square Error, Mean absolute percent error, Mean Absolute Error, Average Error and Root Mean Square Error respectively. The MATLAB software performs the implementation work. Hyderabad city based vehicular pollution data is taken and analyzed to predict environmental pollution based on vehicles. For the analysis of the dataset, the number of pollutants is considered and the Air Quality Index (AQI) is also determined. The proposed hybrid method is tested and evaluated with the existing methods like Artificial Neural Networks (ANN) and Fuzzy Logic Controllers (FLC). In this study, the proposed HWTO‐ANFIS model is utilized to predict vehicular pollution in the last years such as 2018, 2019, 2020 and 2021, respectively. The proposed method's accuracy is 7%–12% greater than that of other conventional techniques. However, the proposed method offers superior detection performances than other state‐of‐art methods.