Two goals were pursued concurrently during this research. The first goal was to develop a performance-driven, steady state, hybrid electric vehicle (HEV) software design tool that would provide design information to the University of Idaho FutureTruck 2000 Suburban. The second goal of this research effort was to develop logic-based, computer algorithms that could be used to outperform numerical solvers currently available.
The HEV design software, SmartHEV, is a flexible and robust model of steady state HEV operation. The power flow through the vehicle components is modeled using the road load power equation. The components that are modeled include the battery pack, alternator, alternative power unit (APU), electric motor, transmission, differential, and wheels. SmartHEV integrates a user-friendly interface with logic-based algorithms that allow the selection of combinations of known and unknown variables. This flexibility in selecting known and unknown variables is a unique feature of SmartHEV. Known variables can also be used to step through a range of vehicle operation in order to calculate optimum performance levels. Parametric results can be plotted and compared against other design configurations. Through the process of selecting known and unknown variables, a user can gain tremendous insight into the relationships of the vehicle components and the variables that describe them.
Several logic-based algorithms were developed and implemented in SmartHEV to improve convergence success and rate of convergence for general systems of equations. The algorithms developed include:
Best Solution Path
The equation Rewriter algebraically manipulates variables within equations to eliminate potential solving problems. The VarSelect routine guarantees that the user specify a non-singular set of unknown variables; the routine also provides tremendous flexibility and insight during the selection of known and unknown variables. The Solution Path and Best Solution Path algorithms determine the best strategy for solving a system of equations; this often results in faster computational speeds when compared to traditional solving strategies. The Swap-and-Solve routine determines alternative solution strategies that can be used to find a valid solution when convergence problems exist. These algorithms are integrated in a Windows environment that manages known and unknown variables and their solutions, making variable identification and selection easier for the user. The engineering design analyst will find that using these algorithms will reduce computational time and improve convergence success when solving linear and nonlinear systems of equations.