Multiobjective Overtaking Maneuver Planning for Autonomous Ground Vehicles.

Area of Science:

• Robotics
• Artificial Intelligence
• Control Systems

Background:

• Generating constrained autonomous vehicle overtaking trajectories is complex due to practical and environmental limitations.
• Optimizing multiple conflicting objectives for irregularly placed obstacles further complicates trajectory generation.

Purpose of the Study:

• To propose a novel swarm intelligence-based algorithm for multiobjective optimal overtaking trajectory generation in autonomous ground vehicles.
• To address challenges in parameter selection by introducing a fuzzy adaptive strategy for dynamic balancing of exploration and exploitation.

Main Methods:

• Developed a swarm intelligence algorithm to solve a multiobjective optimal control model.
• Optimized maneuver time, trajectory smoothness, and vehicle visibility under mission-dependent constraints.
• Integrated a fuzzy adaptive strategy to enhance parameter selection and exploration-exploitation balance.

Main Results:

• Simulation studies validated the effectiveness of the proposed fuzzy adaptive multiobjective method.
• The algorithm successfully produced high-quality Pareto-optimal overtaking trajectories for autonomous ground vehicles.
• Compared to state-of-the-art methods, the proposed strategy generated a more widespread set of optimal solutions.

Conclusions:

• The novel swarm intelligence algorithm with a fuzzy adaptive strategy is effective for generating multiobjective optimal overtaking trajectories.
• The method enhances the ability to explore tradeoffs between objectives, leading to superior trajectory generation.
• This approach offers a robust solution for complex autonomous vehicle navigation challenges.