Adaptive EMS for Dual Power Network vehicles

This paper presents an Adaptive Predictive Control (APC) for Energy Management in order to improve fuel economy in vehicles equipped with a Dual Power Network. This approach can be applied to hybridised and electrical vehicles and provides a locally optimized solution. This is applied at component-level for the Generator Control and in a state-discretized form for the Energy Management system. A practical solution has been the target pursued throughout this research. In this context the term “practical” refers a solution that is executable online as well as easily reconfigurable. The approach proposed is implemented in a vehicle simulator developed within the EE-VERT project. Benefit analysis is referred to ARTEMIS-based real-life driving cycles where a clear distinction between urban and extra-urban cycles is met. Overall an improvement of 60% in average for the Generator Control is attained by replacing the Proportional Integral (PI) controller by an APC. Moreover the spikes in the generator current are completely eradicated. Further, at system level APC achieves quantifiable improvement in fuel economy up to 8.5% (regarding State-of-the-Art strategy) with a maximum depletion of the HV battery of 1.5% against the 22% of Baseline approach (utilising the maximum possible regenerative energy).