Modeling the functional capabilities of the bat wing hair sensor network: effective sensor distributions from optimal sensing.

内容摘要

Effective control of flapping flight can benefit from rapid feedback about wing deformation and aerodynamic loading, yet the functional role of wing-embedded mechanosensors remains poorly understood. In bats, dense arrays of sensory hairs span the wing membranes, but the principles governing their spatial distribution are unknown. We examine the hypothesis that these hairs provide strain-based information sufficient to estimate aerodynamic forces, and that their distribution reflects an evolutionary pressure on sensory performance. Simulations that employ an abstract, simple wing and sinusoidal kinematics produce useful and informative scaling relations: they demonstrate that a fixed relative accuracy of aerodynamic forces requires more sensors for larger wings of a given aspect ratio, and higher sensor density for wings with higher aspect ratios, assuming constant wing area. Using empirical wing kinematics from wind-tunnel flights of Cynopterus brachyotis, the lesser short-nosed fruit bat, we construct a hybrid simulation combining a spring-mass model of the patagia with a blade-element aerodynamic model. We adapt sparse sensor optimization methods from control theory to identify minimal strain-sensing precision distributions that support accurate linear decoding of aerodynamic forces (implicitly minimizing the number of sensory receptors for a given overall accuracy). Optimal solutions concentrate precision near wing boundaries, the propatagium, and the distal part of the handwing. This pattern is qualitatively similar to the sensory hair distribution observed in diverse bat species. Taken together, these results support a role for strain-based proprioception in bat wings that can contribute meaningfully to flight performance. The natural mechanosensation system of the bat wing suggests design principles for bio-inspired flapping-wing sensing systems.