A study on improving the methodology for Torque modeling using a virtual engine model

This study aims to develop an engine torque prediction model using virtual engine simulation data. Accurate torque prediction is essential for minimizing shift shock and ensuring consistent driving performance, particularly in hybrid vehicles where smooth transitions between electric motors and internal combustion engines are necessary. The Engine Control Unit (ECU) uses a physics-based torque prediction model, requiring ignition timing swing data for precise calibration. The virtual engine model, based on 1D gas dynamics, was calibrated using real engine data obtained from a small number of main operating points. The simulation data obtained from the virtual engine model showed a good correlation with the experimental data. By combining large-scale simulation data with limited experimental data, we effectively calibrated the torque prediction model in ECU and confirmed that the calibration results met the development goals. This study demonstrates the potential for efficient engine development method using virtual engine simulations, and we anticipate even better results in the future with more precise EGR correlation.