Optimal control with Kane Mechanics applied to a hybrid power split transmission

摘要

The following work describes a method based on the Kane Mechanics to derive the state space equation for modelling and controlling a mechatronic system such as a hybrid transmission. The hybrid powertrain components, especially the dynamic behaviour for controlling a power split transmission are described and modelled in an object oriented simulation environment. The Kane Mechanics are utilized to build up a state space model of the hybrid transmission. This is utilized to design a Kalman Observer and an Optimal Controller of the type Linear Quadratic Gaussian Integrator (LQGI) for the input speed control. As a benchmark control algorithm a modified PI controller with filter elements is designed that are tuned to compensate the eigenfrequencies of the system. The algorithms are tested and compared within the object oriented simulation environment and implemented into amicrocontroller. Dynamometer and vehicle results show the performance of both algorithms under different conditions. As a shifting quality measure, the Vibration Dose Value (VDV) is used.