无人机精准作业决策方法研究

发布时间：2018-05-28 14:11

本文选题：无人机作业 + 决策规则　；参考：《中国民航大学》2017年硕士论文

【摘要】：由于无人机具有独特的灵活性和优越性,在现代农业、商业以及军事战争中得到了广泛应用,常被用在复杂的、不确定的环境中执行任务。本文主要研究无人机精准作业决策方法中涉及到的决策规则的建立问题,决策规则建立的优劣性是影响无人机作业精准度的关键因素。首先,设计了面向任务的无人机任务顶层体系结构。为了提高各个子系统的自主精准完成任务的能力,全面地选择可能影响无人机决策的因素,分析无人机系统的动力学模型及控制系统结构,设计了任务顶层的体系结构;然后对基于粗糙集理论的无人机作业决策做出了介绍,为接下来的研究奠定了基础。其次,依据粗糙集理论的思想,设计了基于控制系统要素思想的空间属性设计规则;分析了任务空间中属性值的值域覆盖范围对无人机作业的影响,并对连续属性值进行了等间距离散化;然后利用基于粗糙集属性集依赖度的方法求解属性约简,获得去除冗余属性后的决策规则。最后通过无人机的飞行仿真实验,验证依据属性选取规则,以及值域范围的选取是合理和有效的。再次,研究了无人机作业区段的离散化等级程度,提出了基于高斯分布函数的属性离散化等级的设计方法;然后依据精准作业条件属性的值域作业样本分布密度,设计了基于离散分布密度作业区段的属性精细离散化方法,建立了精细离散化后的决策表;最后通过飞行仿真实验验证了离散化精细程度对无人机作业精准度的影响。最后,基于已有的无人机硬件平台,进行了无人机的实际飞行实验。实验结果证明了属性设计规则的可行性;与等间距的离散化方法相比,设计的非等间距精细离散化方法提高了决策的精准度。
[Abstract]:Unmanned aerial vehicles (UAVs) are widely used in modern agriculture, commerce and military warfare because of their unique flexibility and superiority. They are often used in complex and uncertain environments to carry out missions. This paper mainly studies the establishment of the decision rules involved in the decision method of UAV precision operation. The quality of the decision rules is the key factor that affects the operational accuracy of UAV. Firstly, the top layer architecture of UAV mission is designed. In order to improve the ability of each subsystem to complete the task independently and accurately, the factors that may affect the UAV decision-making are selected comprehensively, the dynamics model and control system structure of the UAV system are analyzed, and the top-level structure of the mission is designed. Then the rough set theory based UAV operation decision making is introduced, which lays a foundation for the following research. Secondly, according to the idea of rough set theory, the spatial attribute design rules based on the elements of control system are designed, and the influence of the range coverage of attribute value in mission space on UAV operation is analyzed. The continuous attribute value is dispersed by equal distance, and then the attribute reduction is solved by using the method based on the dependency degree of the attribute set of rough set, and the decision rules after removing redundant attributes are obtained. Finally, the flight simulation experiment of UAV is carried out to verify the rationality and effectiveness of the rules of attribute selection and the selection of range. Thirdly, the degree of discretization of UAV operation area is studied, and the design method of attribute discretization level based on Gao Si distribution function is proposed. The method of attribute fine discretization based on discrete distribution density operation area is designed, and the decision table after fine discretization is established. Finally, the influence of discretization precision on the precision of UAV is verified by flight simulation. Finally, based on the existing UAV hardware platform, the actual flight experiment of UAV is carried out. The experimental results prove the feasibility of attribute design rules, and compared with the equal-spaced discretization method, the non-equal-spacing fine discretization method improves the precision of decision-making.
【学位授予单位】：中国民航大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：V279;TP18

【参考文献】