The Need

Electrified and hybrid vehicle powertrains are increasingly complex, integrating mechanical, electrical, thermal, and emissions subsystems with both continuous and discrete decision variables. Existing energy management and co‑optimization approaches typically rely on simplified models, sequential tuning, or dynamic programming methods that do not scale computationally to realistic systems. Industry lacks a practical framework that can holistically optimize fuel, energy, and emissions performance across full vehicle architectures while remaining computationally tractable for design, calibration, and advanced control use cases.

The Technology

OSU engineers have developed the PS3 framework, a novel, decomposition‑based mixed‑integer optimal control methodology for electrified powertrains. It systematically decomposes a large, multi‑objective vehicle optimization problem into coordinated sub‑problems that jointly optimize design parameters and supervisory control actions. PS3 enables simultaneous optimization of continuous and discrete variables across realistic, high‑fidelity powertrain models, delivering near‑optimal solutions at significantly reduced computational cost.

Commercial Applications

• Supervisory energy management software for hybrid and electric vehicles
• Virtual powertrain design, sizing, and calibration platforms
• Emissions‑aware control strategies for regulated commercial vehicles
• Benchmarking and validation tools for advanced powertrain architectures

Benefits/Advantages

• Holistic optimization: Simultaneously considers powertrain design and real‑time control across all major subsystems
• Scalable mixed‑integer capability: Handles large numbers of continuous and discrete variables beyond traditional DP methods
• Proven performance gains: Demonstrated reductions in fuel consumption, energy demand, and NOx emissions versus baselines
• Architecture‑agnostic: Applicable to parallel hybrids, series hybrids, and range‑extended electric vehicles