Parallel testing of vehicle intelligence via virtual-real interaction

Area of Science:

• Robotics and Control Systems
• Artificial Intelligence
• Automotive Engineering

Background:

• Autonomous vehicle development requires rigorous testing.
• Current testing methods may not adequately challenge advanced systems.
• Accelerated evaluation is crucial for rapid innovation.

Purpose of the Study:

• To introduce a self-driven closed-loop parallel testing system.
• To enhance the complexity and efficiency of autonomous vehicle testing.
• To accelerate the evaluation and development lifecycle of autonomous vehicles.

Main Methods:

• Implementation of a closed-loop control architecture.
• Development of parallel testing capabilities.
• Integration of advanced algorithms for test scenario generation.

Main Results:

• The system successfully executes more challenging test scenarios.
• Demonstrated acceleration in the evaluation phase of autonomous vehicles.
• Facilitated faster iteration cycles for system improvement.

Conclusions:

• The self-driven closed-loop parallel testing system is effective.
• It significantly speeds up the development and evaluation of autonomous vehicles.
• This approach represents a key advancement in testing autonomous driving technology.