SHARP: a hybrid metaheuristic approach for intelligent robotic path planning

Area of Science:

• Robotics
• Artificial Intelligence
• Optimization Algorithms

Background:

• Autonomous navigation requires efficient and safe path planning.
• Existing methods face challenges in complex static and dynamic environments.

Purpose of the Study:

• To present SHARP, a novel hybrid metaheuristic framework for robotic path planning.
• To improve trajectory feasibility, smoothness, and efficiency in autonomous navigation.

Main Methods:

• SHARP combines Particle Swarm Optimization, Sine Cosine search, and Nelder-Mead simplex refinement.
• Two multi-criteria decision layers (Priority-PSN, No-Preference-PSN) were introduced.
• Cubic-spline interpolation was used for trajectory smoothing.

Main Results:

• SHARP consistently produced shorter collision-free paths than other algorithms in static environments.
• The hybrid approach improved path quality in cluttered maps, with a trade-off in computational cost.
• SHARP achieved a high success rate in dynamic replanning, despite increased latency.

Conclusions:

• SHARP offers a practical, adaptable optimization-based framework for intelligent robotic path planning.
• The framework effectively addresses static and dynamic constraints in autonomous navigation.
• Further research may explore latency reduction for real-time applications.