Human driver behaviors significantly affect the control performance and energy efficiency of a powertrain system. This paper investigates the optimization methodology for intelligent control of an electric powertrain adaptive to driver’s intentions and operation patterns. Plant model, driver intention representations, system requirements, and optimization goals are studied. A cyber-physical system (CPS) based framework is proposed for system co-design optimization considering driver intentions. The CPS based performance exploration methodology is presented. Simulation-based system optimizations are carried out. Simulation results show that the overall performance of the electric powertrain is improved with respect to torsional dynamics and energy efficiency, validating the feasibility and effectiveness of the proposed optimization methodology.