DANIEL, T.*; HEDRICK, T; Univ. of Washington, Seattle; Univ. of Washington, Seattle: Inverse analysis of flight control of Hawkmoths

The three dimensional flight paths of insects are the result of complex temporal patterns of muscle activation which, coupled to the exoskeleton, create particular kinematic patterns of wing motion. These kinematic patterns result in aerodynamic forces that drive the animal through a particular spatial path. Most prior research has focused on how such kinematics lead to the forces that underlie the observed motions. However, using only the observed kinematics may not reveal the range of wing motions that might lead to similar � or even identical � body trajectories. The possibility that many kinematic patterns can result in a given flight trajectory has not been considered in any formal sense and may be important to understanding the diversity of motor patterns observed in freely behaving animals. We approached this issue as an inverse problem, asking what kinematics could give rise to the forces required to follow a predetermined flight path. We use genetic algorithms to evolve a suite of control parameters that define wing kinematics and thus flight kinetics in a simulated hawkmoth. These parameters include three angular amplitudes of wing motion (sweep, elevation and pitch), their mean values and relative phases. We find that there are indeed many possible kinematic patterns that can be used to hover or to track a moving flower, a subset of which approximately correspond to those observed in freely flying hawkmoths. These simulations also show that the diversity of feasible control parameters is sensitive to the speed with which controls may change and the number of controls employed. Reducing the number of controls reduces the number of available kinematic patterns while increasing the temporal variation in the controls