Development of Multi-mode Biomimetic Robotic Fish Based on Central Pattern Generator

Abstract

This paper presents the construction and motion control of a biologically inspired, multi-mode biomimetic robotic fish capable of three-dimensional locomotion. The mechanical configuration and the control system are described. The rhythmic movements of swimming are driven by the biological neural networks, called the central pattern generators (CPGs). The CPGs are modeled as nonlinear oscillators for joints and inter-joint coordination is achieved by altering the connection weights between joints. The CPG-based online gait generation method allows elegant and smooth transitions between swimming gaits, which result in multi-mode swimming to implement more lifelike locomotion. Several swimming modes can be obtained to mimic diverse actions of the real fish in nature, or designed according to special assignments by changing CPG parameters. The up-and-down motion can be implemented by adjusting the rotating angle of pectoral fins. The depth of the robotic fish is controlled utilizing a proportional-integral-derivative (PID) control algorithm according to the feedback of depth information measured by the sonar sensor installed at the bottom of the fish. The experimental results confirm the effectiveness of the control methods.