Co-evolution of sensory system and signal processing for optimal wing shape control

Abstract

This paper demonstrates the applicability of evolutionary computation methods to co-evolve a sensor morphology and a suitable control structure to optimally adjust a virtual adaptive wing structure. In contrast to approaches in which the structure of a sensor configuration is fixed early in the design stages, we target the simultaneous generation of information acquisition and information processing based on the optimization of a target function. We consider two aspects as main advantages. First the ability to generate optimal environmental sensors in the sense that the control structure can optimally utilize the information provided and secondly the abdication of detailed prior knowledge about the problem at hand. In this work we investigate the expected high correlation between the sensor morphology and the signal processing structures as well the quantity and quality of the information gathered from the environment.