Experimental trajectory optimization of a flapping fin propulsor using an evolutionary strategy

The experimental optimization of bio-inspired flapping fin trajectories are demonstrated for
potential applications as a side or a rear propulsor of an autonomous underwater vehicle. The
trajectories are scored based upon their difference from a force set-point and upon their efficiency
and are parameterized by 10 variables inspired by fish swimming. The flapping fin is a generic
rectangular rigid flat plate with a tapered edge. Optimization occurs as follows. First, a
generation of trajectories is created. Second, the trajectories are executed by a spherical parallel
manipulator, during which the forces are acquired. Third, the trajectories are scored and a new
generation of trajectories is created using the covariance matrix adaptive evolutionary strategy.
This loop repeats ad-infinitum until the search converges. Within the first set of searches, two
trajectories for optimal side-force generation are found, one is fully three-dimensional while the
other is artificially constr...

Source: Bioinspiration and Biomimetics - November 29, 2018 Category: Science Authors: Nathan Martin and Morteza Gharib Source Type: research