Contributions to Passenger and Commercial Hybrid Electric Vehicle Energy Management Control

Abstract

Passenger and commercial vehicle powertrain electrification are both a technical and a marketing challenge. A common hurdle experienced by the industry and the consumer is limited driving range and a relatively high cost of purchase. Hybrid electric vehicles present a promising solution to meet the ever so stringent fuel economy or energy consumption requirements, as well as comparable driving range to the conventional gasoline vehicles. Though, a challenge with hybrid electric vehicles is the efficient utilization of the available energy to reduce emissions and pitch it as a cost-effective solution. An experimental research vehicle based on the 2016 Chevrolet Camaro platform is re-built as a Plug-in Series Hybrid Electric or Extended Range Electric Vehicle. In this dissertation, mathematical modeling and optimization methods are used to develop a power loss model of this experimental research vehicle. The model is able to estimate vehicle energy consumption within a coefficient of variation of 1.9\%-7.0\%. The model serves as a foundation for the development of a novel energy management optimal control algorithm termed as Distance Constrained - Adaptive Real Time Dynamic Programming (DC-ARTDP). Evaluation of the algorithm over different drive conditions shows an improvement of 9.8\% in the overall energy consumption of the vehicle while meeting required driving range. The algorithm is also able to provide an optimal energy consumption trajectory under powertrain system fault scenarios and meet the required range demand. The algorithms functionality is evaluated against a predictive energy management approach using Model Predictive Control. The novel algorithm improves overall energy consumption of the experimental research vehicle by 4.25\% relative to the model predictive control approach. To understand the impact of electrified powertrain architectures in the context of commercial vehicles, a two-truck and three-truck platoon power loss modeling environment is developed. The impact of platooning on heterogeneous powertrain architecture trucks with and without the proposed novel algorithm is evaluated. Implementation of this algorithm on a Series Parallel hybrid electric truck shows improvements of $\approx$ 8\% for individual runs. Higher improvement is observed for the Series Parallel hybrid electric truck in a lead position comparative to tail position for two and three-truck platoons.