Modeling and control of a hybrid electric drivetrain for optimum fuel economy, performance and driveability

Automotive manufacturers have been striving for decades to produce vehicles which satisfy customers' requirements at minimum cost. Many of their concerns are on fuel economy, road performance and driveability. Improving fuel economy is both a political concern of alleviating dependency on foreign fuel and a customer preference of reducing vehicle operating cost. Consumers also expect vehicles to provide satisfactory performance with desirable driving comfort. Improvements on all these aspects may contribute to lower emissions as well if the vehicle is designed and controlled properly.; A hybrid electric vehicle (HEV) is one of the most promising alternatives to a conventional engine-powered vehicle which satisfies increasing customer requirements mentioned above. However, how much the hybrid vehicle is better than the conventional one depends heavily on its control strategy. The involvement of the electric machine for HEV traction offers the possibility to provide the total tractive force in different ways. Investigations indicate that how to allocate the total tractive force between the engine and the electric machine has significant influences on vehicle fuel economy, performance and driveability. Therefore, designing an optimal control strategy which considers all three criteria is of great interest.; Model based control design requires control oriented models and the complexity of these models are determined by their applications. The control oriented models need to be sufficient to evaluate the control criteria and easy enough for control strategy development. Vehicle fuel economy and performance are related with power allocation in the steady state while human perceptible driveability issues are in the frequency range of a few hertz. Since the control strategy is developed in two steps (finding the solution for the best fuel economy first and then taking driveability into account), two models, i.e., the quasi-static model and the low-frequency dynamic model are built for each step in the control design. Simulation results demonstrate that these two models are effective to capture the main behaviors of the vehicle and to evaluate fuel economy, performance and driveability respectively.; Defining objective metrics for vehicle fuel economy, performance and driveability is also very important. Evaluations in both simulations and real vehicles require objective and quantitative measures. Subjective and descriptive metrics cannot be easily implemented in simulations and evaluations vary with changing time or evaluators. (Abstract shortened by UMI.)...