| In this dissertation, some control strategies for a particular hybrid electric vehicle using a reasonable logic switch, fuzzy logic and neuro-fuzzy are developed to minimize energy use. A first requirement before the controller could be developed was a suitable simulation environment. In order to investigate the control of a hybrid drive train, it is necessary to develop computer models of the components that make up such a vehicle. The modeling approaches taken use both reverse and forward dynamic models, and embody an understanding of real component operation.; The First objective of this work is to investigate the electrically peaking control concept, then, to design a controller which insures steady state torque command tracking while offering the user some choices of parameters to optimize. The second objective of this study is to investigate the use of fuzzy logic techniques for the control of a hybrid electric vehicle drive train. The practical requirement is a controller capable of blending the operation of the Internal Conbustion Engine (ICE) and the electric motor.; The concept of an adaptive fuzzy network (AFN) and its combination of gradient descent and least square estimate learning procedure are discussed in detail. Simulation results of the logic switch controller are used as the training data set for the AFN design. Simulation results illustrating the proposed approach are demonstrated over a standard test driving cycle. |
