Data-driven based eco-driving control for plug-in hybrid electric vehicles

With the development of connected and automated vehicles, eco-driving control is reckoned to generate unprecedented potential on energy-saving in electrified powertrain. In this paper, a data-driven based strategy with efficient computation capacity proposed for plug-in hybrid electric vehicles achieve approximate optimal energy economy. An hierarchical scheme designed mitigate massive computational cost during velocity optimization powertrain control. A consumption model an battery current are respectively constructed via two neural networks served as critic system optimization. Furthermore, network-based dynamic programming exploited optimize vehicle by merging models Bellman optimality principle. The simulation results demonstrate that method can remarkably improve fuel economy up 16.7% complicated driving conditions, compared conventional sequential methods. drastically efficiency slight sacrifice economy, optimum benchmark.