A novel predictive semi-physical feed-forward turbocharging system transient control strategy based on mean-value turbocharger model

Variable geometry turbine is a technology that has been proven on diesel engines. However, despite the potential to further improve gasoline engines’ fuel economy and transient response using variable geometry turbine, controlling the variable geometry turbine during transients is challenging due to its highly non-linear behaviours especially on gasoline applications. After comparing three potential turbocharger transient control strategies, the one that predicts the turbine performances for a range of possible variable geometry turbine settings in advance was developed and validated using a high-fidelity engine model. The proposed control strategy is able to capture the complex transient behaviours and achieve the optimum variable geometry turbine trajectories. This improved the turbocharger response time by more than 14% compared with a conventional proportional–integral–derivative controller, which cannot achieve target turbocharge speed in all cases. Furthermore, the calibration effort required can be significantly reduced, offering significant benefits for powertrain developers. It is expected that the structure of this transient control strategy can also be applied to complex air-path systems.