Research on Path Planning Algorithm of Driverless Ferry Vehicles Combining Improved A* and DWA.

Area of Science:

• Robotics and Autonomous Systems
• Artificial Intelligence
• Navigation and Control

Background:

• Global path planning algorithms struggle with unknown dynamic and static obstacles.
• Local planning algorithms often face local optimization issues in large-scale environments.
• Driverless ferry vehicles require robust and efficient path planning for safe operation.

Purpose of the Study:

• To develop an integrated path planning algorithm combining A* and Dynamic Window Approach (DWA) for driverless ferry vehicles.
• To enhance obstacle avoidance capabilities, particularly for unknown dynamic and static obstacles.
• To address the limitations of traditional algorithms, including local optimization and path inefficiency.

Main Methods:

• Improved A* algorithm with enhanced heuristic function (vector angle cosine) and optimized search neighborhood.
• Path smoothing using cubic quasi-uniform B-spline curves to reduce turning points.
• Fuzzy control theory integrated into DWA for dynamic adjustment of evaluation function weights.
• Fusion of improved A* for global planning and improved DWA for real-time local obstacle avoidance.

Main Results:

• The integrated algorithm successfully avoids unknown dynamic and static obstacles in real-time.
• The system achieves global optimal paths while ensuring efficient local navigation.
• Simulation results demonstrate the algorithm's effectiveness and adaptability across various environment maps.

Conclusions:

• The proposed A*-DWA fusion algorithm provides an efficient and robust solution for path planning in driverless ferry vehicles.
• The enhancements to both A* and DWA effectively mitigate their respective limitations.
• The algorithm offers real-time obstacle avoidance and adaptability, crucial for autonomous navigation.