Combining Optimization and Simulation for Next-Generation Off-Road Vehicle E/E Architectural Design.

Area of Science:

• Automotive Engineering
• Electrical and Electronic Systems
• Systems Design

Background:

• Off-road vehicles face unique challenges integrating Advanced Driver Assistance Systems (ADASs) and autonomous driving.
• Current electrical-electronic (E/E) architectures struggle with data demands, necessitating zonal architectures and Drive-by-Wire technologies.
• Transitioning to new communication standards is crucial for future vehicle functionalities.

Purpose of the Study:

• Propose an automated methodology for designing next-generation off-road vehicle E/E architectures.
• Address the limitations of current E/E architectures in supporting advanced vehicle functionalities.
• Optimize E/E architectures for performance, efficiency, and cost.

Main Methods:

• Utilized a MILP-based optimizer to identify potential E/E architectural solutions.
• Employed discrete event simulation to validate the feasibility of proposed architectures.
• Applied an ascent-based gradient method to refine optimizer constraints for convergence.

Main Results:

• Evaluated architectural solutions based on latency, jitter, and network load.
• Conducted a Pareto analysis considering power consumption, cost, and system weight.
• Demonstrated the effectiveness of the automated methodology in generating optimized E/E architectures.

Conclusions:

• The proposed automated methodology effectively designs next-generation off-road vehicle E/E architectures.
• The approach balances performance metrics with economic and physical constraints.
• This methodology provides a robust framework for future automotive E/E system development.