Evolutionary multiobjective design of a flexible caudal fin for robotic fish

Area of Science:

• Robotics
• Bio-inspired Engineering
• Computational Optimization

Background:

• Robotic fish utilize body or fin deformation for locomotion.
• Flexible materials offer enhanced swimming performance but pose design challenges.
• Robotic fish have applications in environmental monitoring and biological studies.

Purpose of the Study:

• To develop an evolutionary multiobjective optimization approach for robotic fish design.
• To optimize morphological and control parameters for a flexible caudal fin robotic fish.
• To balance competing objectives of swimming speed and power efficiency.

Main Methods:

• Employing the NSGA-II algorithm for evolutionary multiobjective optimization.
• Investigating morphological and control parameters for flexible fin designs.
• Experimental validation using multi-material 3D-printed fins with varying stiffness and size.

Main Results:

• Identified optimal parameter values for enhanced swimming speed and reduced power consumption.
• Demonstrated the effectiveness of the evolutionary approach in design optimization.
• Validated evolved fin designs through physical experiments with a robotic fish.

Conclusions:

• The proposed evolutionary multiobjective optimization effectively balances speed and power efficiency in robotic fish.
• Flexible caudal fins designed with this method improve robotic fish performance.
• This approach provides a robust framework for designing advanced bio-inspired robots.