Force Optimization of Elongated Undulating Fin Robot Using Improved PSO-Based CPG

Area of Science:

• Robotics
• Control Systems
• Biomimetics

Background:

• Biorobotic fish offer advanced underwater capabilities, including speed and maneuverability.
• Locomotion control in these robots is crucial for performance.
• Existing models require optimization for enhanced propulsive efficiency.

Purpose of the Study:

• To investigate an enhanced central pattern generator (CPG) model for biorobotic fish locomotion.
• To implement differential particle swarm optimization (D-PSO) for CPG parameter tuning.
• To improve swimming speed and propulsive performance of an undulating fin robot.

Main Methods:

• Developed a CPG network with sixteen coupled Hopf oscillators for gait generation.
• Introduced differential particle swarm optimization (D-PSO) to optimize CPG parameters.
• Compared D-PSO with traditional PSO and genetic algorithm (GA) for parameter tuning.
• Tested the D-PSO-based CPG on a physical undulating fin robot.

Main Results:

• The D-PSO-optimized CPG network increased thrust force, leading to faster swimming speeds.
• The D-PSO method demonstrated superiority over traditional PSO and GA in CPG parameter tuning.
• The optimized undulating fin robot showed a 5.92% average increase in propulsive force compared to the standard CPG model.

Conclusions:

• The D-PSO-based CPG model effectively enhances the propulsive performance of biorobotic fish.
• This optimization approach leads to significant improvements in swimming speed and efficiency.
• The study validates the effectiveness of D-PSO for controlling complex biorobotic systems.