Prediction of the level of air pollution using adaptive neuro-fuzzy inference system

Air pollution is a severe environmental issue that has garnered international attention. Air pollution forecasting is critical for daily health monitoring and government decision-making. Current study methodologies, on the other hand, have been unable to adequately separate the geographical features of air pollution concentration data, resulting in fast changes in long-term accuracy and air quality. Many techniques of air pollution detection are available which developed by researchers. However, these techniques are not achieved efficient accuracy. Hence, in this paperis develop a hybrid adaptive neuro-fuzzy inference system (ANFIS) and a recurrent neural network (RNN). The RNN technique used to calculate pollution load has a direct influence on quantifying pollution’s impact on air quality and on the overall assessment results. As a result, the weight of each assessment criteria must be determined in a convoluted and thorough manner. This Fuzzy is engaged in air pollution assessments for CO, NO2, O3, PM2.5 and PM10. In five types of hybrid model investigations, the suggested system’s similarity was utilized to separate the three most acceptable climatic factors from six typical climatic characteristics (atmospheric pressure, relative humidity, air temperature, wind speed, wind direction, and total precipitation). The selected form, which included humidity, wind speed, and wind direction, provided high forecast accuracy. For each level of CO, NO2, O3, PM2.5 and PM10, we also presented the accuracy, sensitivity, specificity, accuracy, susceptibility, and F1 scores to assess the ANFIS-RNN prediction outcomes. On Python platforms, design and deploy unique air pollution coding systems. Simultaneously, comparative investigations revealed that ANFIS-RNN outperforms ANN and RNN samples. The collected findings indicate the efficacy of air pollution forecast analysis for an effective air quality forecast.