Front‑Loaded Virtualization Drives Advanced AWD‑Like Motion Control for PATAC’s RWID ATV
Replacing Mechanical Differentials with Virtualized Control Strategies Using Simulink
“From our early 0.1 platform to the current 3.1 model library architecture, these standardized toolchains allow projects to reuse tools and processes, improving development efficiency and quality while lowering onboarding barriers.”
Key Outcomes
- Simulink was integrated with PATAC and third-party tools to enable virtualization pre-development in the vehicle motion control field and build a three-layer software architecture.
- Using Simulink, PATAC achieved virtualization in the RWID ATEV to support features such as intelligent escape from difficult terrain, drive redundancy, minimal-turning radius, circular steering, and parallel parking.
- Vehicle development projects attained stable motion and controllability under motor fault conditions.
Vehicle development projects are becoming increasingly complex due to shorter development cycles, delayed system-level integration, testing difficulties, and the competitive need for advanced features. To better address these challenges, the Pan Asia Technical Automotive Center (PATAC), a Shanghai-based joint venture between General Motors and SAIC Motor, is utilizing Simulink® to develop a vehicle model that integrates virtual and physical approaches.
The PATAC team implemented front-loaded virtualization in the vehicle motion control domain on their first extended-range Advanced Technology Engineering Vehicle (ATEV), equipped with rear-wheel independent drive (RWID). Subsystem-level targets defined software functions and control algorithms. Simulink enabled driving scenarios testing, such as full-throttle straight-line acceleration, cornering, hill climbing, and split-friction surfaces, leading to better tracking and agile steering. Simulating motor failure cases boosted vehicle controllability.
The simulation results were fed back into improving control strategies. The team developed a control strategy to replace mechanical differential locks with electronic systems. Dynamic torque distribution at micro and macro levels, drivetrain control, active damping, and Kalman filters helped to achieve torque smoothness. At the final stage, they implemented cloud-based monitoring and a standardized, reusable algorithm library to integrate closed-loop simulations with real-world testing.
