PSO-GRNN–Based Adaptive Online Vehicle Velocity Prediction Strategy Considering Traffic Information

In order to increase the online vehicle velocity prediction (VVP) strategy’s forecast performance, an adaptive VVP strategy considering traffic signals is presented for multiple scenarios. Initially, the algorithm of a general regressive neural network (GRNN) paired with data sets of the ego-vehicle, the vehicle in front, and traffic lights is used in traffic scenarios, which increasingly improves the prediction accuracy. To ameliorate the robustness of the algorithm, then the strategy is optimized by particle swarm optimization (PSO) and k-fold cross-validation to find the optimal parameters of GRNN in real-time, which constructs an adaptive online PSO-GRNN VVP strategy with multi-information fusion to adapt with different operating situations. To verify the proposed strategy, traffic scenarios are established inside the co-simulation environment. The adaptive online PSO-GRNN VVP strategy is then deployed to a variety of simulated scenarios to test its efficacy under various operating situations. Finally, the simulation results reveal that in urban and highway scenarios, the prediction accuracy is separately increased by 31.3% and 48.3% when compared to the traditional GRNN VVP strategy with fixed parameters utilizing only the historical ego-vehicle velocity data set.