Smart Traffic Eco Powertrain
The real driving emissions of vehicles with an internal combustion engine often deviate from the values measured during homologation. One of the main reasons for this, are the significantly higher dynamic components in real driving scenarios compared with those in the currently mandatory drive cycles. Fast load changes lead to transient states in the combustion process which the engine control unit must continuously compensate by control interventions. Despite complex exhaust aftertreatment systems, the lack of knowledge of the driving profile is a major challenge. An intelligent control of the battery state of charge and a suitable recuperation strategy are decisively for the drivetrain efficiency of vehicles with electric and hybrid drive. Therefore, the most accurate possible knowledge of the driving profile and a predictive control are crucial.
However, if in the future the control unit is provided with the information of an imminent load change, for example when starting at a green light via the V2X communication, a control intervention which takes effect before the actual load increase, could precondition the control variables for a reduced pollutant emission for combustion engines. In electric and hybrid vehicles, the same information could be used for control interventions in the battery, drive and hybrid control systems.
To estimate the potential of these ideas, the Online Calibration Tool has already been further developed in the DBU-funded research project "NET-ECU - Networked Engine Control". Both, interventions in the vehicle control system and data exchange with novel information sources are possible. At the same time, the required computing capacity was made available for new algorithms that reduce emissions with the help of V2X information. Investigations have shown that a control-based, preventive strategy for controlling boost pressure in simulations reduces soot emissions by up to 7% and NOx emissions by up to 1%.
The follow-up project "STEP - Smart Traffic Eco Powertrain" aims to extend the promising approaches of the previous project and to test these in real traffic situations. The previously developed OCT will be extended by additional interfaces for further sensors so that vehicles which are not equipped with V2X can also be detected by using radar, lidar and camera. In addition, the software will be extended so that additional traffic information sources can be used. The previously developed algorithms for reducing emissions will be supplemented in such a way, that energy requirements will also be reduced. These algorithms will then be transferred to electrified drive trains.