Pattern Recognition for Robotic Fish Swimming Gaits Based on Artificial Lateral Line System and Subtractive Clustering Algorithms

The complicated and changeable underwater environment increases the difficulty of pattern recognition for robotic fish swimming gaits. Aiming at this question, environment sensing and pattern recognition using an artificial lateral system are investigated in this work. Imitating lateral line of real fish in nature, a novel artificial lateral line system for robotic fish is designed in this paper. Based on this novel system, the pressure information around robotic fish can be sensed when robotic fish swims in different gaits, so the feature points can be extracted from the pressure information. And then, based on the feature points, a subtractive clustering algorithm is used to recognize the swimming gaits of robotic fish. So the pattern state of robotic fish can be obtained, which provides a basis for the quick control of robotic fish in water. Finally, a validation experiment is conducted with freely swimming robotic fish. The validity of this novel system is demonstrated. And the feasibility and accuracy of subtractive clustering algorithm used in pattern recognition for robotic fish is verified too