Monitoring in-Vehicle Air Quality in Real Time Through IoT and Enhanced Deep Learning Models

Globally, air pollution (AP) is among the most pressing environmental crises. With human health in mind, numerous researchers have focused on these issues. Predictive data on air quality is one of the most effective means of alerting the public to the dangers of air pollution and encouraging them to take preventative measures. One of the most difficult environmental problems in many large cities is air pollution. Preprocessing, feature selection, and model training are the three stages that make up the process. Data scaling, pollution comparison, and status fixing are all parts of preprocessing. The process of feature selection makes use of Q-learning based bee swarm optimization. The model was trained using an A-CNN-LSTM-XGBoost. With an average accuracy of $\mathbf{9 4. 6 1 \%}$, the suggested method outperforms LSTM and A-CNN.