Multi-Strategy Improved POA for Global Optimization Problems and 3D UAV Path Planning

Area of Science:

• Robotics and Automation
• Artificial Intelligence
• Aerospace Engineering

Background:

• Drone technology is crucial for smart manufacturing and the low-altitude economy.
• Path planning is a core challenge impacting drone efficiency and safety.
• Existing algorithms often require extensive data or get stuck in local optima.

Purpose of the Study:

• Introduce a novel algorithm for Unmanned Aerial Vehicle (UAV) path planning.
• Address limitations of current path planning methods, such as local optima and data dependency.
• Enhance the efficiency, safety, and robustness of drone missions.

Main Methods:

• Developed a Multi-strategy Enhanced Pelican Optimization Algorithm (MIPOA).
• Implemented a hybrid initialization combining low-discrepancy sequences and heuristic refinement.
• Incorporated subgroup mean-guided updating and a random reinitialization boundary mechanism.

Main Results:

• MIPOA demonstrated superior optimization capability compared to eleven benchmark metaheuristics on the CEC2017 test suite.
• Statistical analyses confirmed the algorithm's enhanced performance.
• Successfully applied MIPOA to 3D UAV path planning in realistic threat scenarios.

Conclusions:

• MIPOA offers a robust and adaptable solution for UAV path planning.
• The algorithm effectively overcomes challenges like local optima and premature convergence.
• MIPOA ensures successful mission completion in complex, dynamic environments.