QQ客服 官方微博 反馈留言

题名	小型无人直升机动力学建模、辨识及鲁棒飞行控制系统设计与优化研究
作者	王晓东
学位类别	工学博士
答辩日期	2010-05-26
授予单位	中国科学院研究生院
授予地点	中国科学院自动化研究所
导师	谭民 ; 赵晓光
关键词	小型无人直升机 数学建模 预报误差法 飞行控制系统 H∞回路成形 多目标优化 MOGA-II small-scale unmanned helicopters mathematical modeling prediction error method flight control system H∞ loop shaping multi-objective optimizatio MOGA-II
其他题名	Modeling, Identification and Robust Flight Control System Design and Optimization of Small-Scale Unmanned Helicopters
学位专业	控制理论与控制工程
中文摘要	小型无人直升机是近年来发展迅速的一种航空机器人，具有可垂直起降、悬停、任意方向巡航飞行、定点旋转、高机动性和高敏捷性等优点，在军用和民用方面具有重要的应用价值和广阔的应用前景。研究表明，小型无人直升机是典型的多输入多输出欠驱动系统，具有不稳定、高非线性、强耦合、全飞行包线内模态及特性多变等特点。飞行控制系统是无人直升机系统的核心，是实现无人直升机自主飞行的关键。 本文针对小型无人直升机的特点对其自主飞行进行了深入的研究，包括小型无人直升机数学建模、模型辨识、鲁棒飞行控制系统设计与优化等等。本文的主要工作和贡献如下： 1、在研究了直升机飞行理论和空气动力学理论的基础上，应用第一性原理建立了小型无人直升机数学模型，包括用于仿真的非线性模型和用于控制器设计的线性模型。通过对直升机耦合特性的分析，提出了适用于实验辨识的小型无人直升机的横/纵向耦合模型、偏航模型和升降模型。 2、分析了预报误差法(PEM)的原理和辨识过程，并提出了一种应用PEM辨识得到小型无人直升机系统横/纵向通道的MIMO耦合模型和偏航通道、升降通道的SISO模型的方法。通过飞行实验，辨识得到了小型无人直升机各模型的参数。 3、提出了一种H∞回路成形小型无人直升机鲁棒飞行控制器设计方法，采用该方法分别设计了小型无人直升机的俯仰/滚转姿态控制器、升降控制器和偏航控制器。将所设计的控制器与PID控制器进行对比，仿真结果表明，H∞回路成形设计方法在确保鲁棒稳定性的同时，能有效降低横、纵向通道间的耦合，并具有良好的指令跟踪能力和时域响应性能。 4、 针对H∞回路成形设计方法的设计关键——权重函数的选取问题，提出了一种采用多目标遗传算法MOGA-II的权重函数选取和优化框架。仿真结果表明，该方法不仅可以减少调整权重函数的次数，缩短控制器的设计周期，并且最终得到的控制器不仅具有较强的鲁棒性，而且可以同时满足频域和时域的性能指标要求。 最后，对所开展的工作进行了总结，并指出了需要继续开展的研究工作。本论文开展的研究工作和取得的阶段性成果，为提高小型无人直升机的控制性能进行了有益的探索，将对今后的研究工作起到良好的推进作用。
英文摘要	The small-scale unmanned helicopter, an aerial robot developed in recent years, has high flexibilities and agilities such as vertical taking off and landing, hovering, cruising in arbitrary directions, pirouetting etc. Having such unique merits, the small-scale unmanned helicopter possesses great value and significant potentials in both military and civil applications. Researches manifest that the small-scale unmanned helicopter is a typical multi-input multi-output (MIMO) and under-actuated system, which is inherently unstable, highly nonlinear, strongly coupled, and together with its mode and dynamics varied in full flight envelope. Considering the complicated dynamic characteristics of the small-scale unmanned helicopter system, the development of its flight control system (FCS) is difficult, and therefore the key to achieve its autonomous flight.. In this thesis, a study is made on the characteristics of autonomous flight of small-scale unmanned helicopters, which involves mathematical modeling, model identification, robust flight control system design and optimization, etc. The main contributions of this thesis include following issues: 1. A study is made on the helicopter's flight theory and aerodynamic theory. On the basis of the theory studied and the first-principle modeling method, we build the mathematical models, including a nonlinear one for simulation, and a linear one for controller design, for the small-scale unmanned helicopter. With the analysis of the coupling characteristics of the small-scale helicopter, we build models suitable for the experiment of system identification. These models include the lateral and longitudinal coupling model, yaw model, and heave model. 2. The principle and identification procedure of prediction error method (PEM) is studied. Using the PEM identification method, we propose an approach to identify an MIMO coupling model for lateral and longitudinal channel, and SISO models for yaw channel and heave channel, respectively. With flight experiments, we obtain the parameters of such models of a small-scale unmanned helicopter. 3. Based on the H∞ loop shaping design procedure (LSDP), a robust controller design method is proposed for the flight control of small-scale unmanned helicopters. With this method, we design the controllers for the control of pitch/roll attitude, yaw angular rate, heave velocity, respectively. The comparison study from simulation with PID controllers and the controllers obtained wi...
语种	中文
其他标识符	200518014628008
内容类型	学位论文
源URL	[http://ir.ia.ac.cn/handle/173211/6244]
专题	毕业生_博士学位论文
推荐引用方式 GB/T 7714	王晓东. 小型无人直升机动力学建模、辨识及鲁棒飞行控制系统设计与优化研究[D]. 中国科学院自动化研究所. 中国科学院研究生院. 2010.

个性服务

查看访问统计

相关权益政策

暂无数据

收藏/分享

除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。