Hybrid Electric Powertrain Design and Control with Planetary Gear Sets for Performance and Fuel Economy

Abstract

Planetary gear sets (PGs) play a key role in hybrid electric vehicle (HEV) design by enabling a variety of unique architectures using a limited number of powertrain components. Leveraging the capability of this mechanical device, this study introduces an automated design process for PG-based HEV systems focusing on both fuel economy and performance, while also deriving the necessary analysis and synthesis tools. First, the design process generates all possible modes in an HEV design with a given set of powertrain components. The data structure and the derivation method of speed and torque relationships of each mode enable an exhaustive search of the large design space that grew with all the component topology and PG gear ratio combinations. Second, all powertrain types realizable with a given set of components are mathematically shown, and each feasible mode is classified under one of these powertrain types. Third, computationally efficient linear programming solvers suitable for vector operations are developed for each powertrain type to assess the forward- and backward-speed gradeability, long-hauling torque, and acceleration time of each mode for all PG gear ratio combinations. Fourth, the combination of modes that meets the performance requirements, along with the number and location of clutches that make these mode transitions possible, are identified. As a result, each potent mode combination, the clutches necessary for the mode transition, and the auxiliary modes established through all clutch state combinations constitute a design that meets the performance criteria. Last, the fuel economy improvement potential of each design is evaluated using an algorithm that approximates dynamic programming optimization. The results show that light-duty truck performance requirements can be met by many two-PG HEV designs without sacrificing fuel economy if the right analysis and synthesis techniques for exploring the entire design space are developed. In addition to the design process, the feasibility of mode transitions and the effect of mode transitions on the fuel economy simulation results are investigated. For this purpose, the dynamics of mode transition is analyzed, and control algorithms achieving the transitions without interrupting the desired vehicle torque are developed. Then, these analysis and synthesis techniques are automated so that they can be integrated into the fuel economy simulation algorithm. The simulation results reveal that some mode transitions have a negative effect on fuel economy and the assumption of mode transition feasibility at any operating point is not valid.