A bionic intelligent method combining evolutionary game theory with particle swarm optimization for UAV 3D path...

Area of Science:

• Robotics
• Artificial Intelligence
• Optimization Algorithms

Background:

• Path planning for Unmanned Aerial Vehicles (UAVs) in complex 3D environments is crucial for motion control systems.
• The NP-hard nature of UAV path planning presents significant challenges in generating high-quality paths.
• Existing methods struggle with efficiency and optimality in obstacle-rich environments.

Purpose of the Study:

• To propose an improved self-adaptive particle swarm optimization (ISAPSO) algorithm for UAV path planning.
• To enhance the exploration and exploitation balance in optimization algorithms.
• To develop an ISAPSO-based path planner for optimal 3D path generation in complex environments.

Main Methods:

• Integration of standard Particle Swarm Optimization (PSO) 2011 with Evolutionary Game Theory (EGT).
• Development of a novel self-adaptive parameter updating strategy using EGT's evolutionary stable strategy and hyperbolic tangent function.
• Implementation of a self-adaptive constraint handling technology for efficient constraint management in path planning.

Main Results:

• The proposed ISAPSO algorithm demonstrated superior performance against six state-of-the-art evolutionary algorithms in benchmark tests.
• ISAPSO achieved a 90% confidence level in outperforming competitors on 20 test functions.
• The ISAPSO-based path planner significantly outperformed counterparts in path optimality across various scenarios.

Conclusions:

• The ISAPSO algorithm offers a robust and effective solution for UAV path planning in complex 3D environments.
• The novel self-adaptive strategies enhance optimization capabilities, leading to superior path quality.
• This method presents a vital alternative for advancing UAV motion control and autonomous navigation systems.