Scenario-Based Simulation for Advanced Driver Assistance Systems Validation

Advanced Driver Assistance Systems (ADAS) and intelligent vehicle control technologies require rigorous validation before deployment. While real-world testing offers realism, it remains expensive, time-consuming, and often limited by regulatory constraints or unsuitable for evaluating early-stage control strategies or edge-case behaviors. Virtual simulation provides a safe, configurable, and repeatable environment for functional testing, particularly for systems that rely on rule-based activation and real-time control response. This thesis presents a co-simulation framework for function-level vehicle control validation. The platform integrates MATLAB/Simulink for control algorithm design with RoadRunner, which implements road scenario generation and traffic modeling. Real-time interaction is achieved through standardized interface blocks, allowing the control system to dynamically respond to lane geometry, lead vehicle behaviors, and other driving events defined in simulation. To evaluate system performance, a range of representative test scenarios has been designed. These include typical highway and urban driving cases involving speed regulation, stopping behavior, and lateral tracking. System behavior is assessed based on control smoothness, rule compliance, and activation thresholds. In addition, the platform supports the integration of custom controllers and real-world GPS-based map data, allowing adaptation to tasks such as trajectory tracking, motion planning, and performance-oriented driving. The proposed framework introduces a modular and simulation-driven strategy for testing ADAS functions and demonstrates its potential as a practical foundation for academic research, early-stage controller development, and regulation-oriented intelligent vehicle validation.