Advanced control concepts for a parallel hybrid powertrain with infinitely variable transmission

• Main Author: Cacciatori, Enrico
• Other Authors: Vaughan, N. D.
• Published: Cranfield University 2006
• Subjects: 629.2446
• Online Access: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.443335

Poweitrain systems of increasing complexity are being introduced by automotive manufacturers in order to reduce carbon emissions into the atmosphere: hybrid electric vehicles and continuously variable transmissions represent effective contributions to achieve the emission reduction target. The increased complexity calls for more sophisticated control strategies to be developed; different energy management approaches have been investigated in the past, in most cases without considering driveability requirements. Those strategies relying on the knowledge of future driving conditions cannot be deployed in a real-time controller and are only used to investigate patterns of optimal behaviour. This Thesis investigates two energy management strategies for an innovative parallel hybrid powertrain concept with innately variable transmission. This was developed as part of a government funded research project aiming at demonstrating the potential fuel economy benefit of such driveline configuration. Both strategies have a common architecture and rely on a common scheme to control the transient vehicle response; this was experimentally calibrated in order to provide improved driveability levels with respect to the conventional non hybrid powertrain deploying the same transmission concept. This control scheme and its calibration are maintained across the two energy management strategies so that consistent vehicle behaviour is achieved and the cost of driveability in terms of energy usage is preserved. The first energy management strategy was heuristically formulated to maximise operation of the single powertrain components in conditions of maximum efficiency. Optimal design techniques were adopted for the calibration of the corresponding rule set. The second strategy formulation was based on the analysis of the simulation results obtained from a dynamic programming model; regression analysis techniques were used to provide the necessary knowledge base required for the control rules formulation and calibration. ln both cases engineering intuition is required for the interpretation of the simulation results and for the individuation of patterns of behaviour. The hybrid powertrain provides consistent fuel economy improvements with respect to the equivalent non hybrid powertrain with innately variable transmission. A driveability appraisal was conducted and the subjective ratings showed an improved overall driveability level of the hybrid powertrain. Despite producing different control and state trajectories, both strategies provide similar fuel economy figures across a set of legislative drive cycles thus demonstrating that both approaches effectively exploit the hardware limits of the powertrain plant.