Traffic Condition Prediction for Urban and Highway Scenarios using Artificial Intelligence

A major study topic, particularly in machine learning and artificial intelligence, has resulted from the prediction of traffic congestion in recent years (AI). During the last few decade, this study field has greatly increased due to the emergence of huge data from stationary sensor or probe vehicle data. Congestion forecasting requires analysing a wide range of traffic variables, especially for short-term planning. Predictions of future traffic congestion are based mostly on analyses of past patterns. Using traffic data like taxi trip records, urban traffic statistics, and junction counts, we propose a new method for calculating traffic flows. The suggested method makes use of CGAN and GCN, and we enhanced Unet to realise a key component of the generator. The interaction between junctions with surveillance and intersections without monitoring is then captured using floating taxis that are spread out over the whole city. The CGAN architecture may modify the weights and improve the inference capability to provide an accurate traffic status under the present circumstances. The results of the experiment demonstrate that our technique is more accurate at estimating traffic volume than other methods.