Rapid Prototyping Hybrid Powertrain Research Platform


Student: Yu Wang, Ali Sadighi


Project Introduction:
As one of the most promising approaches for reducing automotive fuel consumption, hybrid powertrain has inspired extensive research efforts on hybrid architecture design and energy management optimization. However, the complexity and cost of constructing or physically modifying a hybrid powertrain system seriously slow down the experimental investigation of its complicated dynamics, so as to limit the improvement on precise control and optimization of the hybrid powertrain. To provide an accurate and flexible hybrid powertrain emulation tool for hybrid powertrain development, a rapid prototyping hybrid powertrain research platform. This well greatly expedite the research various hybrid powertrain architectures and control methodologies, without actually building the complete physical system.


Project Outline:
To construct the hybrid research platform, three levels of research work are being conducted, including: 1) Physical level: design and control of the transient hydraulic dynamometer, which will provides the hardware ingredient for the proposed platform; 2) System level: design, control and operation coordination of the multi-level hybrid powertrain platform, which will focus on the control synthesis for the various subsystems in the complicated platform; 3) Advanced algorithm level: design and experimental validation of the comprehensive-optimal hybrid energy management strategy, which is both an application of the hybrid powertrain research platform, but also a further investigation in the core area of  the hybrid powertrain research.

First, as the basic hardware ingredient in the proposed research platform, a hydrostatic dynamometer with an electronically controlled load sensing mechanism is designed and implemented. To control the dynamometer to emulate the real-world engine speed/torque profiles, systematic nonlinear tracking control strategies (nonlinear model-based inversion control, and state feedback control via feedback linearization) are investigated and implemented. Less than 3-5% tracking error for engine speed control is obtained in experiments. 

Second, based on the well-controlled transient dynamometer, a multiple-level hybrid powertrain simulation and control system is designed for realizing the hardware-in-the-loop hybrid powertrain emulation. Particularly, corresponding to the high/middle/low level in the simulation and control system, an adaptive driver model, a DP based energy optimization strategy, a decoupled based hybrid torque controller and a dynamometer torque controller are integrated and coordinated.

Third, based on the research platform, the further investigation will be carried out in the core area of the hybrid powertrain technology, i.e., the energy management strategy development.  A real-time hybrid energy management strategy is being investigated and its associated performance (fuel efficiency, emissions and drivability) will be characterized based on the hybrid powertrain research platform.



Fig. 1 Hardware overview of the hybrid powertrain research platform


Fig. 2 Schematic diagram of the hybrid powertrain research platform


Fig. 3 Control system architecture and hardware-in-the-loop modeling


Figure 4 Hybrid operation emulation experiments with the Highway Fuel Economy Test (HWFET)


  1. Wang, Y. and Sun, Z., “Hybrid Powertrain Control with a Rapid Prototyping Research Platform”, Proceedings of the 2011 American Control Conference, San Francisco, CA, pp.997-1002, June, 2001.
  2. Wang, Y., Sun, Z. and Stelson, K., “Modeling, Control and Experimental Validation of a Transient Hydrostatic Dynamometer”, in press, IEEE Transactions on Control Systems Technology, 2011.