Extremum Seeking Algorithm to Optimize Fuel Injection in a Hydraulic Linear Engine

A Hydraulic Linear Engine (HLE) concept designed by the United States Environmental Protection Agency incorporates a crank for improved robustness but transmits power hydraulically Previous work has shown that each cylinder behaves differently requiring that injector timing and duration commands be managed individually in order to balance and optimize engine performance. The high-fidelity simulation summarized in this paper uses physics-based models and established correlations to predict closed-loop engine response to the reported control structures. The proposed control strategy utilizes an adaptive, cylinder-balancing algorithm to adjust fuel quantity for each cylinder and an extremum seeking algorithm to adjust fuel injection timing for each cylinder. Modeled results show an appreciable decrease in total fuel consumption using the improved injection timing.