A Path Tracking Method and Validation Experiment for Robotic Manta

The mission-oriented control of bionic robot is one of the important problems in the field of underwater robotics, in which the study of path tracking is essential. To cope with the study, the paper proposes a model-free path-tracking method for robotic manta combining line of sight (LOS) method, TS fuzzy neural network and artificial central pattern generators (CPG) network. Firstly, the prototype and coordinate system definition of the robotic fish are introduced, and the multi-motor distributed pectoral fins are driven using the improved artificial CPG network. Then, the control of the pitch angle and heading angle is accomplished by the TS fuzzy neural network. Accordingly, the LOS-based 3-D path-tracking guidance law is designed. Finally, to validate the methodology, experiments are carried out in 3-D space. The experiment data proved that the proposed path tracking method can successfully track the desired path. The method is clear in structure and simple to implement, and it is hoped that it provides new support for the task execution of robot fish in complex environments.