Control-relevant parameter estimation application to a model-based PHEV power management system

Explicit model predictive control approach is a promising approach to fulfill automotive real-time controls requirements. A key factor in the performance and real-time capabilities of a predictive model-based controller is the accuracy of the control-oriented model. The control-oriented model should capture the essential dynamics of the real plant and be adequately simple to make the controller implementable on a commercial hardware with limited memory and computational capabilities. In this study, control-relevant parameter estimation is used to find a control-oriented model for a real-time predictive power management system for a plug-in hybrid powertrain. Simulations, which are conducted using an equation-based model of the powertrain, demonstrate a significant improvement of the power management system performance by improving the control-oriented model with no effect on real-time capabilities of the controller.