Toward Turning Performance Optimization of a Multi-Flexible Robotic Fish

Over prolonged evolution, natural fish possess exceptional maneuverability. For underwater robotic fish, maneuverability is a crucial performance metric during locomotion. This paper proposes a turning control optimization framework for a multi-flexible-joint bionic robotic fish. Firstly, three distinct turning strategies are devised based on kinematic analysis of the flexible robotic fish’s turning motion. Experimental results indicate that the control strategies need to be adjusted with respect to motion frequency to achieve optimal performance for the flexible robotic fish. Furthermore, an optimization problem incorporating dynamic constraints and cost functions of the flexible robotic fish is constructed. Based on a Constrained Iterative Linear Quadratic Regulator (CILQR), a turning performance optimization method is designed. Subsequently, optimal control strategies under both stationary and motion states are derived, effectively enhancing the turning performance of the flexible robotic fish. Finally, simulation and experimental results validate the effectiveness of the designed method. The developed flexible robotic fish could achieve a maximum swimming speed of 1.63 BL/s (body lengths per second) and a maximum average turning speed of 130.5°/s, offering guidance for the practical application of flexible robotic fish in aquatic environments. Note to Practitioners—This paper was motivated by the problem of improving maneuverability in underwater robotic systems. While traditional underwater vehicles often struggle with agile movement in complex environments, this paper suggests a solution inspired by nature. The key practical contribution is a turning control optimization framework that enhances maneuverability without requiring complex mechanical redesigns. We’ve developed three distinct turning strategies that can be selected and adjusted based on operational requirements. Our research shows that motion frequency significantly impacts turning performance—a critical insight for field deployment. The optimization method we propose can be implemented in existing control systems with minimal computational overhead, making it accessible for various underwater applications. This approach could benefit underwater inspection, environmental monitoring, and search operations where navigation in confined or obstacle-rich environments is required.