Crank-Angle Resolved Real-Time Engine Modelling: A Seamless Transfer from Concept Design to HiL Testing 2018-01-1245

Virtual system integration and testing using hardware-in-the-loop (HiL) simulation enables front-loading of development tasks, provides a safer and reliable testing environment and reduces prototype hardware costs. One of the greatest challenges to overcome when performing HiL simulations is assuring a high model accuracy under stringent real-time requirements with acceptable development effort. This article represents a novel solution by deriving the plant model for HiL directly from the existing detailed models from the component layout phase using co-simulation methodology. It provides an effective and efficient model implementation and validation process followed by detailed quantitative analysis of the test results referred to the engine test bench measurements.

For virtual calibration purpose, a detailed one-dimensional (1D) GT-POWER model for a state-of-the-art turbocharged diesel engine with exhaust gas recirculation (EGR) is simplified and transformed to a HiL platform connected to an engine control unit (ECU). The engine model remains semi-physical and crank angle resolved. The major pressure pulsations within the system are well captured, which is mandatory for the determination of volumetric efficiency, turbocharger operation and EGR distribution. A predictive combustion model based on injection profiles is implemented for modelling of the indicated engine efficiency and the exhaust gas temperature. After detailed investigations on steady-state and transient model performance in an offline environment, the model is integrated into the HiL testing platform. The coupling of the model to the ECU interface has been implemented using the co-simulation approach on FEV’s xMOD platform. The simulation results of the integrated HiL system, including the engine thermodynamics and the controller behaviours, have been validated with measurement data from engine test bench, and the real-time capability of the model has been proven. The work has demonstrated the capability and advantages of a seamless transfer from component design to system integration and testing within a combustion engine development process.