Modelling and control of a parallel through-the-road plug-in hybrid vehicle
With an increasing number of passenger vehicles, automotive emissions has become a major challenge. Among other pollutants, green house gas (GHG) emissions form a majority of the exhaust coming out of the tail-pipe and are blamed for the rising temperature of the earth. The net carbon emissions due to running a lightduty passenger vehicle are studied in the present thesis. Hybrid electric vehicles have been proven to have lower green house gas emissions than conventional vehicles as electricity is cleaner than fossil fuels. Using a parallel-through-the-road hybrid architecture vehicle, different power management strategies are studied. To begin with, a detailed model of the vehicle is developed based on dynamometer testing. The power management algorithms developed are implemented on these models instead of the real vehicle. Dynamic programming has been used to find optimal GHG emissions for the test vehicle. The dynamic programming solution is found to result in a 19% improvement in GHG emissions (fuel consumption in charge-sustaining mode) and is also used as a benchmark for other power management approaches such as equivalent consumption management strategy and proportional state-of-charge algorithm. As dynamic programming cannot be implemented by itself, an approach is proposed to determine trends from the optimal solution, and implement it on the software models developed initially. The other techniques, although not as good as dynamic programming, are found to give almost similar GHG emission benefits.
Meckl, Purdue University.
Off-Campus Purdue Users:
To access this dissertation, please log in to our