Agent- AI Based Framework for Green Route: Enhancing Travel Using Best Pathfinding With SSFS and SE-TDT-SVM Through AQI-Aware Routing

Abstract

Air pollution poses significant health risks in urban environments. It necessitates accurate Air Quality Index (AQI) prediction and route planning. In urban environments, it is becoming increasingly important. The importance of managing and mitigating air quality has been highlighted in a various recent study. This work presents a novel approach integrating real-time air quality data with Agent based AI techniques. It predicts AQI levels and recommend low-pollution travel routes in major Indian cities. It proposes a SALP Swarm based Feature Selection (SSFS) algorithm to identify key pollutant indicators. This also proposed a Stacked Ensemble of Tuned Decision Tree and Support Vector Machine (SE-TDT-SVM) for AQI prediction. AQI-Aware Routing (AAR) algorithm optimizes path selection based on both air quality and distance metrics. The SE-TDT-SVM model achieved a prediction accuracy of 94.5% with an error rate of 5.5% on the feature-selected dataset. The AAR algorithm successfully generated multiple path options, balancing travel distance and AQI exposure. This integrated approach offers a robust solution for urban air quality management and health-conscious route planning. The potential applications in smart city initiatives and public health interventions.