Path planning and smoothing of mobile robot based on improved artificial fish swarm algorithm

Area of Science:

• Robotics
• Artificial Intelligence
• Computational Geometry

Background:

• Traditional artificial fish swarm algorithms face challenges in path planning accuracy and convergence.
• Mobile robot path planning requires smooth, kinematically feasible trajectories.

Purpose of the Study:

• To develop an improved artificial fish swarm algorithm for mobile robot path planning and smoothing.
• To enhance path accuracy, convergence, and ensure continuity in orientation and curvature.

Main Methods:

• Integration of Dijkstra's algorithm for feasible solutions and step sizes.
• Introduction of a dynamic feedback horizon and adaptive step size for improved convergence.
• Application of Bessel curve theory for path smoothing.

Main Results:

• Achieved 100% path planning accuracy in simulations.
• Ensured the shortest average path length in the tested grid environment.
• Generated paths continuous in orientation and curvature, meeting robot kinematic requirements.

Conclusions:

• The improved artificial fish swarm algorithm effectively addresses limitations of traditional methods.
• The proposed approach provides accurate, smooth, and kinematically feasible paths for mobile robots.