Perception Sensors Based Fusion System and Method for Controlling Vehicle

Taxi demand prediction has lately piqued academic attention due to its vast potential for use in expansive intelligent transportation systems. On the other hand, the majority of earlier methods merely took into account forecasting taxi demand in origin locations, omitting to simulate the particular circumstances of customers traveling to a destination. This research study presume that it is inefficient to pre-allocate taxis to each location based just on origin demand. This research study offers a significant and fascinating challenge in this work termed taxi origin-destination demand prediction, which tries to estimate future taxi demand between all area pairings. Its main challenges are figuring out how efficiently a variety of contextual data are gathered in order to understand demand trends. In this study, origin and destination views were among the classification tasks where a unique Deep Neural Network (DNN) with Deep learning-based models outperform conventional machine learning techniques. Comprehensive testing and analysis on a large dataset clearly demonstrate that, for forecasting taxi origin-destination demand, our DNN beats other evaluated approaches.