Genetic Algorithms in Search, Optimization and Machine Learning
Genetic Algorithms in Search, Optimization and Machine Learning
Construction and Modelling of a Carangiform Robotic Fish
The Sixth International Symposium on Experimental Robotics VI
Hi-index | 0.02 |
This work presents an approach for maneuvering and controlling a biomimetic autonomous underwater vehicle (BAUV). The BAUV swims forward by oscillating its body and caudal fin. It turns by bending its body and caudal fin toward the intended direction of motion. A body-spline function is specified by a set of parameters. Genetic algorithms are then used to find the values of the parameter by evaluating a fitness function over several swimming trials in a water tank. The fitness function is defined as the ratio of the kinetic energy of the forward motion to the required driving power of the joint motors. A control law that uses the oscillating frequency to control the forward speed, and applies a body-spline offset parameter to control the yawing rate is proposed. Moving averages of swimming speeds and heading angles are utilized as feedback signals to control the forward speed and heading angle of the BAUV. The effectiveness of the control algorithm is experimentally confirmed.