ECS-9810242 Freudenberg Advanced technology powertrains are being proposed to satisfy demands for lower emissions and higher fuel economy without sacrificing drivability. The novel actuators added to meet these goals typically interact with more than one powertrain subsystem. Current control calibration strategies are based upon multiple single control loops, and cannot compensate for these interactions. Instead, sophisticated multivariable control will be needed to achieve the performance potential of advanced engines. The goal of the proposed research is to develop and apply system theoretic concepts to the problem of controller design for highly interacting advanced powertrains. In particular, we address the problem of selecting a multivariable controller architecture that best meets the tradeoff between system performance and control complexity. The Principal Investigator will work with engineers at Ford who have access to models of advanced powertrains and dynamometer test facilities with which to evaluate proposed control algorithms. Automotive powertrains are a challenging application field in which methods of systems and control are needed to solve important technological problems.
