小型ロボットで進化した人工免疫系行動の大規模プラットフォームへの転移
The Transfer of Evolved Artificial Immune System Behaviours between Small and Large Scale Robotic Platforms
シミュレーション上で小型ロボット(epuck)に進化させた行動を、形状やセンサ構成の異なる大型ロボット(Pioneer P3-DX)へ転移し、適応型人工免疫系制御により物体追跡タスクを遂行できることを示した。
著者: Amanda Whitbrook, Uwe Aickelin, Jonathan M. Garibaldi
分類: cs.NE, cs.RO
原文アブストラクト
This paper demonstrates that a set of behaviours evolved in simulation on a miniature robot (epuck) can be transferred to a much larger scale platform (a virtual Pioneer P3-DX) that also differs in shape, sensor type, sensor configuration and programming interface. The chosen architecture uses a reinforcement learning-assisted genetic algorithm to evolve the epuck behaviours, which are encoded as a genetic sequence. This sequence is then used by the Pioneers as part of an adaptive, idiotypic artificial immune system (AIS) control architecture. Testing in three different simulated worlds shows that the Pioneer can use these behaviours to navigate and solve object-tracking tasks successfully, as long as its adaptive AIS mechanism is in place.
