大工至善|大学至真分享 http://blog.sciencenet.cn/u/lcj2212916

博文

[转载]【武器系统】【2018】非线性自适应导弹自动驾驶仪设计

已有 1313 次阅读 2020-10-14 21:25 |系统分类:科研笔记|文章来源:转载

本文为德国慕尼黑技术大学(作者:Florian Ulrich Peter)的博士论文,共254页。

 

防空导弹(如地对空导弹或空对空导弹)具有高敏捷性和高速度的特点。在防空场景中,导弹由追踪器引导朝向目标飞行,以达到与目标轨迹相交(直接命中)或最小化拦截点偏离的目的。对目标的拦截可分为三个场景阶段:推进段、中段和结束段在最后一个阶段,导弹系统需要满足高要求的轨迹,以使拦截点处目标与导弹之间的距离最小。根据导弹制导单元发出的这些轨迹,自动驾驶仪为导弹的执行器部分生成指令。因此,导弹自动驾驶仪是决定系统闭环性能和跟踪特性的关键元件。在传统的设计方法中,集成控制结构的目的是建立与线性参考动态一致的闭环特性。强制非线性导弹系统在一组大量的操作点上表现出线性、一致的行为,导致闭环性能特性远远落后于导弹的最大物理能力。

 

在本论文中,我们以现代非线性控制方法为基础,发展一种新颖的自动驾驶仪架构,目的是充分发挥导弹在整个作战范围内的性能。自动驾驶仪结构分为三个部分:非线性参考、基线控制律和自适应增强。与传统的、整体的自动驾驶方法不同,本文开发的模块化体系结构允许根据特定系统的性能和鲁棒性要求对每个部件进行独特的分配。

 

针对非线性参考模型和基线控制律的布局,定制了非线性动态逆和反步技术,以适应导弹的动力学特性,满足标称条件下的要求。为了保持这种闭环行为,即使在大的模型偏差情况下,由于参数和传感器的不确定性,在自动驾驶仪中加入了一个基于L1分段常数的级联自适应结构。在整个设计和验证过程中,采用了一个非线性、六自由度的地对空导弹仿真模型,包括真实的气动数据集以及执行器和传感器单元的动态表示。通过线性分析和非线性仿真验证了本文所开发的自动驾驶仪结构相对于线性方法的优越性。

 

所提出的新型自动驾驶仪结构和相应的设计过程不仅限于导弹应用。某些元素、程序和注意事项可以为具有非线性特性的任何高空平台的未来控制设计增加重要价值。

 

Missiles for air defense purposes (e.g.surface-to-air missile or air-to-air missile) are characterized

by their high agility and fast velocities.In anair defense scenario the missile constitutes the pursuer guiding itself towardsthem aneuvering threat(evader) with the purpose to intersect the target’s trajectory(direct-hit) or to minimize the deviation at the point of intercept. Theintercept of a target can be subdivided in three scenario phases: boost,midcourse, and endgame. Within the last phase, the missile system needs to fulfilldemanding trajectories to minimize the distance between target and missile atthe point of intercept. Based on those trajectories, issued from the missile’sguidance unit, the autopilot generate commands for the missile’s actuatorsection. Therefore, the missile autopilot constitutes the key elementdetermining the system’s closed-loop performance and tracking characteristics.In traditional design approaches, integrated control architectures are appliedwith the purpose of setting up the closed-loop characteristics coinciding witha linear reference dynamics. Enforcing the nonlinear missile system to exhibita linear, uniform behavior at a large set of operating points leads toclosed-loop performance characteristics lagging far behind the missile’s maximumphysical capabilities.

Within this thesis, a novel autopilotarchitecture is developed based on modern, nonlinear control methodologies withthe purpose of fully exploiting the missile airframe’s performance capabilitiesacross the entire fight envelope. The autopilot architecture is subdivided in threeelements: a nonlinear reference, a baseline control law, and an adaptiveaugmentation. In contrast to classic, holistic autopilot approaches, the hereindeveloped modular architecture allows an unique assignment of each element withrespect to certain system’s performance and robustness requirements.

For the layout of the nonlinear referencemodel and the baseline control law Nonlinear Dynamic Inversion and Backsteppingtechniques are tailored to match the missile’s dynamical peculiarities and fulfillthe demanding requirements under nominal conditions. To maintain thisclosed-loop behavior even in cases of large model deviations, stemming fromparametric and sensor uncertainties, a cascaded adaptive structure based on L1-Piecewise-Constant is incorporated within the autopilot. A nonlinear, sixdegree of freedom surface-to-air missile simulation model including a realisticaerodynamic data set and dynamic representations of actuator and sensor unitsis used throughout the entire design and verification process. Linear analysisand nonlinear simulations are utilized for proving the superiority of theherein developed autopilot architecture compared to linear methods.

Furthermore, the general validity of thelayout process across the entire fight envelope is demonstrated. The proposednovel autopilot architecture and the corresponding design process are notlimited to missile applications only. Certain elements, procedures, andconsiderations can add significant value in future control design of any aerialplatform exhibiting dominant nonlinear characteristics.

 

 

1. 引言

2. 导弹模型

3. FSD通用地空导弹(FGS-X-03)飞行动力学分析

4. 非线性自动驾驶仪设计与分析的数学背景

5. FGS-X-03模型的飞行控制系统

6. FCS分析与评估

7. 总结、结论与展望

附录坐标框架

附录空气动力学数据集


更多精彩文章请关注公众号:205328s611i1aqxbbgxv19.jpg




https://wap.sciencenet.cn/blog-69686-1254400.html

上一篇:[转载]【计算机科学】【2020.03】非传统领域的深度学习
下一篇:[转载]【电子技术】【2013.08】自学习传感器研究:基于FPGA的ADC前端
收藏 IP: 111.39.173.*| 热度|

0

该博文允许注册用户评论 请点击登录 评论 (0 个评论)

数据加载中...

Archiver|手机版|科学网 ( 京ICP备07017567号-12 )

GMT+8, 2024-3-29 13:58

Powered by ScienceNet.cn

Copyright © 2007- 中国科学报社

返回顶部