A Scalable Approach for Energy Converter Modeling and Supervisory Control Design

Abstract

Increasing demand for hybrid electric vehicle designs requires automated modeling and simulation tools to construct a design space search. Composability and scalability are highly desirable in these simulators to provide design candidates. To achieve effective and efficient automation of the design space search, a scaling approach with moderate complexity and consideration of component nonlinear characteristics is necessary in sizing powertrain components, especially energy converters. The Willans line approach proposed in this paper provides models that are sufficiently accurate but not excessively complicated and permits the derivation of a scalable supervisory control algorithm for overall energy consumption minimization. The scalable models and the control strategy can be adapted to various vehicle drivetrain configurations, providing a scalable and composable simulation tool for vehicle powertrain rapid designing. Scaling results of the Willans line approach are compared with real data and another scaling method in a “forward”, quasi-static simulator.