RI:Collaborative Research: Robust Ornithopter Flight - from Engineering Models to Cooperative Indoor Maneuvers

扑翼飞行为室内环境提供了高机动性。为了实现搜索和室内导航等任务的鲁棒智能，扑翼飞机的机动性将与一种学习方法相结合，这种学习方法对干扰和障碍物的性质做出最小的假设。 我们建议开发算法，扑翼飞机在没有事先地图的情况下，在典型的室内环境中进行感知和导航。 我们的研究将通过全三维动力学仿真、多绳系实验室试验台以及实际室内飞行的扑翼飞机进行验证。这项工作要解决的关键研究问题是：1）改进的扑翼飞机力学和空气动力学2）扑翼飞机鲁棒飞行控制策略3）仅使用简单传感器信息的扑翼飞机协同导航学习算法这项研究将促进对室内飞行的高机动性扑翼飞行器的理解。 通过在一个真实的室内环境中结合从力学到习得行为的研究，我们将测试如何将每个层次的表现整合起来以实现鲁棒智能。我们的项目将为学生提供跨学科的教育，通过力学、传感、控制和学习的结合来实现鲁棒智能。 这项研究可能会导致飞行机器人能够稳健地进入未知和危险的室内环境，可能会使救援人员远离危险。

Flapping flight provides high maneuverability for indoor environments.To achieve robust intelligence for tasks such as search and indoornavigation, the maneuverability of an ornithopter will be combinedwith a learning approach which makes minimal assumptions about thenature of disturbances and obstacles. We propose to developalgorithms for ornithopters to cooperate in sensing and navigation intypical indoor environments without prior maps. Our research will beverified with full three dimensional dynamic simulation, amulti-tethered laboratory test-bed, as well as with actual indoorflying ornithopters.The key research issues to be addressed in this work are:1) improved ornithopter mechanics and aerodynamics2) robust ornithopter flight control strategies3) learning algorithms for cooperative navigationof ornithopters using only simple sensor informationThis research will advance understanding of high maneuverabilityflapping wing vehicles for indoor flight. By combining research fromthe levels of mechanics to learned behavior in a real indoorenvironment, we will test how performance at each level can beintegrated to achieve robust intelligence.Our project will provide interdisciplinary education for students inachieving robust intelligence through the combination of mechanics,sensing, control, and learning. This research can lead to flyingrobots which can robustly enter unknown and hazardous indoorenvironments, potentially keeping rescue workers out of harms way.