Analytical approach for sensitivity analysis of hybrid electric vehicle design

This paper introduces a novel approach for evaluating the sensitivity of an optimal powertrain design for plug-in hybrid electric vehicles. The proposed methodology is based on a reformulation of the original problem as a linear optimization problem from which an analytic solution can be determined. The obtained expression enables the cost function to be evaluated at very low computational cost. The methodology and the results will be introduced on the example of a truck's total cost of ownership optimization for different powertrain sizing. It will be shown that with this approach the results can be obtained 40 times faster than with a bi-level optimization strategy while the average relative difference is less than 0.35%.