Success Story

Hyundai Engine Control Strategy Development

SimuQuest has provided Hyundai Motor Company with a set of tools that enable the rapid development of engine control algorithms. With this new capability, HMC can now develop algorithms in a simulation environment. Once ready for virtual calibration and simulation testing, these algorithms are simply placed into the platform engine control architecture. They can then be tested in simulation in conjunction with the rest of the engine strategy connected to a highly accurate model of the engine and vehicle. Alternatively, the entire software architecture is generated for testing either on the Hardware In the Loop (HIL) system, on an engine dynamometer or in the test vehicle.

Test Bench for Rapid Prototyping and HIL Testing:

The following remarkable results were obtained:

• Engine strategy first tested and calibrated with successive simulations.
• Engine strategy flashed to MPC565 module and tested on the HIL. Results verified to be the same as seen in simulation.
• Module tested in the vehicle. Vehicle started smoothly first time, idled and driven. Results almost identical to those seen on the HIL and in simulation.

List of tools used:

• Matlab/Simulink/Stateflow/RTW Embedded Coder
• SimuQuest Enginuity for accurate engine model
• SimQuest LTI++ for engine control design (P+I, LQG, H∞)
• SimuQuest UniPhi for model architecture and data management
• SimuQuest QuantiPhi MPC56x blockset – device drivers
• SimuQuest OSEK blockset for MPC56x
• SimuQuest custom blockset for Metrowerks TPU powertrain drivers
• dSPACE HIL for real time execution of engine model and module test
• Autonet MPC565 test module
• Metrowerks TurboOSEK
• Metrowerks TPU drivers for powertrain control
• ETAS INCA for HIL and vehicle calibration and data logging

Key Development Steps:

• Enginuity was used to obtain an accurate model of the 4 cylinder engine. Engine mapping data was provided and then used by the Enginuity automatic calibration tool to tune the engine model.
• An engine control strategy was rapidly developed using the UniPhi architecture and data manager.
• Strategy simulation tested in Simulink using Enginuity for the engine model. The entire engine control strategy (fuel, spark, air estimation, start, idle, etc.) was simulation tested and calibrated.
• Alternative controllers for idle speed and closed loop fuel were developed using the Enginuity LTI++ tool (PI, LQG and H∞).
• All lower layer software (OS, CAN, IO drivers) were built in Simulink using the QuantiPhi MPC56x blockset.
• Metrowerks engine specific drivers, for CAM, fuel, spark, etc., were integrated into a Simulink blockset.
• An OSEK blockset was also provided – so that either OSEK OS or conventional engine strategy OS could be implemented.
• Strategy features and lower layer (drivers and OS) were integrated in Simulink enabling code generation of the entire software architecture.
• Software compiled and flashed to the MPC565 test module.
• INCA used for fine tuning of calibration parameters and monitoring of signals.
• Module first tested on a dSPACE HIL (DS1005) running Enginuity and vehicle model in real time.
• Module then tested in the vehicle. Vehicle started and idled smoothly the first time. Data collected verifying close correlation of the vehicle data and simulation results. Switched between P+I, LQG or H∞ controllers in vehicle.