Predictive Hybrid Powertrain Energy Management with Asynchronous Cloud Update

The optimal energy management of a hybrid powertrain has the task to provide the required traction power combining both power sources in the best way. This can be achieved well if the future drive cycle is known/precomputed. However, both speed and traction power requirement may deviate from the expected ones due to many factors, like traffic, weather etc. Against this background, it might be sensible to recompute them whenever needed to keep using the latest future information. Unfortunately, this computation is typically too slow for real time use. In this paper we propose a control structure in which the real time task is solved by a predictive controller which tracks the optimal reference from the cloud, and requests an update of the reference regularly. The update can integrate new information from V2X. This asynchronous operation allows recovering most of the performance of the perfect prediction, while removing tight constraints on the offline computation and copes better with interruptions in communications to the cloud.