Establishing a Model Based Toolchain for ECU Validation

Abstract

This thesis details the establishment of a model-based toolchain for the validation of a motorcycle Engine Control Unit (ECU). The increasing complexity of modern ECUs presents significant challenges for traditional testing methodologies, necessitating more efficient and comprehensive validation approaches. This work addresses these challenges through the development of a Hardware-in-the-Loop (HiL) simulation environment. The primary objective was to create an adaptable platform capable of replicating a motorcycle's operational environment to thoroughly test ECU functionalities. This was achieved by developing detailed component models in MATLAB Simulink, employing a hybrid modelling strategy that combined behavioral, data-driven, and physics-based techniques. These models, representing key components connected to the ECU, were integrated into a main simulation model. This central model was then incorporated into the Vector CANoe software environment for real-time execution on a pre-existing Vector Test System HiL testbench. A structured variant control system was also implemented within Simulink to manage different motorcycle variants with their varying parameters and configurations. The software toolchain was configured to enable consistent model compilation and integration for ECU interaction. The validation of the developed toolchain demonstrated its effectiveness. A detailed four-stage test case, encompassing engine start-up, warm-up, a dynamic drive cycle, and engine cooldown, was simulated as single continuous simulation. This test sequence ran without inducing any Diagnostic Trouble Codes from the connected ECU, confirming that the models stimulated the ECU correctly across the varied operating conditions. The results verified the successful integration of the component models, the functionality of the variant control system, and the overall integrity of the HiL simulation environment. This project successfully established a reliable and adaptable model-based simulation platform. This platform provides an effective alternative to traditional vehiclebased testing methods and can be used for more advanced diagnostic procedures, such as On-Board Diagnostics (OBD)-I and OBD-II testing. The developed toolchain contributes to more efficient and comprehensive ECU validation processes within the motorcycle industry.

Date of Award	2025
Original language	English
Supervisor	Harald Kirchsteiger (Supervisor)