【作者】 魏榛；

【导师】 杨基明；

【作者基本信息】 中国科学技术大学 ， 流体力学， 2010， 博士

【摘要】 昆虫高超的飞行本领一直是仿生研究关注的一个热点。昆虫飞行研究可分为机理研究与应用研究:机理研究通过理论分析,活体观测,模型实验和数值计算等手段来揭示扑翼运动的非定常空气动力学机理;应用研究通过设计和研制人造扑翼机构来模拟昆虫的飞行能力,进而实现扑翼微飞行器等仿生应用。目前机理与应用研究都得到了很好的发展:对扑翼飞行的高升力机理的认识在不断深入和细化,具有实用价值的扑翼微飞行器也在崭露头角。但扑翼研究的探索与实践过程也表明,关于昆虫飞行仍有不少尚未解决的问题需要探讨。本文立足于实验手段,对活体观测,模型实验和扑翼微飞行器研制等问题进行了一系列研究。本文针对活体观测研究中海量昆虫图像的形态学特征自动识别难题提出了一种新的“帧间差别法”。该方法通过巧妙地利用相邻三帧图像间的运算和昆虫内在特点来实现运动特征的自动提取。多组虚拟和真实昆虫运动图像的测试结果表明:这种方法能够较好的适应昆虫这种复杂变形体运动;并能够在翼面透明,昆虫身体有遮挡,高速摄影本底噪声大的环境中顺利地自动提取二值图并分析形态学特征。此方法无需人工干预,不针对特定昆虫,运算量也较小,识别结果可靠。这给昆虫图像形态学特征自动识别提供了一个切实可行的解决方法,极大的节省了昆虫扑翼运动图像的分析时间。研制了一种由舵机驱动的大比例鹰蛾扑翼模型。舵机控制方便,结构紧凑,成本低廉,结合所设计的控制程序使得计算机能够智能、便捷地精确控制该模型实现所需扑翼运动。此模型与鹰蛾在运动学和形态学上都较好地满足相似,能够直接在空气里进行扑翼实验,这更好的模拟了昆虫飞行环境。使用该模型进行的鹰蛾悬停飞行实验结果表明:攻角与挥拍角间的相位差显著地影响气动力的产生,适当的超前相位角有助于获得悬停飞行所需的气动力;扑翼运动曲线本身对气动力的影响也很重要:真实鹰蛾运动曲线比简化方式的运动曲线气动力特性要好,更有利于悬停飞行;鹰蛾悬停飞行中翼面的扭转变形并不显著影响其气动力;过度超前的相位对平均升力的提升很有限且会增加气动功耗。鹰蛾扑翼方式本身在所研究的情况中已经接近最优,这在一定程度上揭示了昆虫飞行的内在机制。针对超微型扑翼机构设计难度大、加工代价高的特点,本文提出并验证了一种新型的“电磁驱动翼面”机构。该微型扑翼机构把驱动装置与翼面合二为一,只需要对薄膜翼面线圈进行控制即可实现扑翼运动。此外,针对平面扑翼机构难实现挥拍角与攻角精确控制的问题,本文设计了“平行曲柄连杆”扑翼机构。通过对该机构进行多参数优化和气动力测量表明,这种扑翼机构能够很好地模拟昆虫扑翼运动,并能提供昆虫自重相当的升力来达到悬停飞行要求。此外本文还将扑翼飞行的高升力优势应用到仿生水下航行器的研制,获得了高推力性能。这拓展了扑翼飞行的研究价值,也有利于更机动灵活的水下航行器开发。更多还原

【Abstract】 The surprising ability of insect flight is always a focus of biomechanics. Insect flight research can be divided into mechanism research and application research: the mechanism research can reveal the unsteady aerodynamics mechanism of flapping flight by theoretical analysis, insect observations, flapping model experiments and computational fluid dynamics; the purpose of the application research is to build flapping micro air vehicle (MAV) through the design and manufacture of artificial flapping mechanism which can simulate insect flight motion. Nowadays, the mechanism research and application research are both well developed: the unsteady high-lift mechanism in insect flight is proved, flapping MAV which has practical value is also developed. However, there still many problems about the insect flight have not been solved yet and need to be explored. In the present work, a series of experimental investigations were caried out for the insect observation, flapping model experiments and Flapping MAV development.A novel method called‘Frame Difference’is proposed to solve the problem that the massive works of extraction from enormous amount of insect images can be treated in an acceptable period of time to obtain the morphological characteristics of insect flight. A noise-suppress operation between three adjacent frames and the internal characteristics of insects are fully used to realize the motion characteristics automatic extraction. The tests on some virtual and real insect image sequences indicated that: this method can well adapt the complex deformable contour of insect flight; and moreover, it still can provide good results even when the insect wings are partly transparency, or the body covers the wings or the high-speed camera has significant noise. The method can work without manual interference, and with no limitation for certain species of insects. Only relatively simple calculation is needed and in most cases it can give robust extraction results. The method is proved to be a feasible solution of massive insect images’morphological characteristics automatic extraction problem, and it can significantly reduce the time cost of the insect images sequence analysis.Considering the normal flapping models’problems that transmission complexity, huge size, lack of mobility and high cost, here a large-scale Hawkmoth flapping model driven by Servos is developed. Servos are very cheap, easy to control and they have compact structure. Using the intelligent software designed for the model control through personal computer, the Hawkmoth model can perform accurate flapping motion as real Hawkmoth. The model has similarity not only on the aerodynamics but also on the morphology. Furthermore, it can directly simulate the real Hawkmoth in the air, which is much closer to the real condition. The results of aerodynamics experiments for hovering Hawkmoth simulation with this model reveal that: the phase lag between stroke angle and angle of attack would have significant influence on the lift generation, and a suitable advance phase angle should be beneficial to obtain hovering lift; the motion curves are also important: real Hawkmoth motion mode can produce more lift than the simplified motion mode, so it is better for hovering flight; the wings’twist deformation along the wingspan wise during the hovering flight can’t result to the remarkable increase of lift; the lift increase from excessive advance phase angle is very limited and it would increase aerodynamic power requirement. All the results show that: real Hawkmoth motion mode has the best aerodynamics performance among our experimental data, and This may explain the miraculous flight ability of Hawkmoth that derived from evolution.In view of the ultra small flapping mechanism research difficulty that hard to design and high cost of manufacture. Here a novel flapping mechanism named‘electromagnetic drive wing’is proposed and justified. The mechanism integrates drive part and wing part into one film coil, and it is very easy to obtain flapping motion when the coil is controlled by current. In order to overcome the shortcoming of plane flapping mechanism that it can’t control the stroke angle and angle of attack all together, a novel flapping mechanism called‘Parallel Crank-Rocker’was successful developed. By the multi-parameters optimization and force measurement experiments, it was shown that the novel mechanism can perfectly simulate the insect flapping motion and can provide sufficient lift for the hovering flight requirement.In addition, some developments of flapping propulsion devices were challenged, in which flapping wing’s high lift capability was referenced for the bionic underwater vehicle design to obtain high thrust. It was demonstrated that the great potential of flapping wing will be beneficial to the development of high performance bionic underwater vehicle.更多还原