• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文会议>Saint Petersburg International Conference on Integrated Navigation Systems; 20060529-31; Saint Petersburg(RU) >EVALUATION OF A NEW APPROACH FOR THE DESIGN OF GENERIC 4D GUIDANCE SYSTEMS
【24h】

EVALUATION OF A NEW APPROACH FOR THE DESIGN OF GENERIC 4D GUIDANCE SYSTEMS

机译：通用4D制导系统设计新方法的评估

获取原文
获取原文并翻译 | 示例
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

Differential flatness, a property of some dynamic systems has made possible the development of new tools to analyze and to control complex nonlinear dynamic systems. Certain outputs presenting the flatness property are of special interest since it becomes possible, in theory by inverting the system dynamics, to derive from their planned trajectory the corresponding control inputs. Some papers have investigated the differential flatness of conventional aircraft full flight dynamics. In the case of air transportation, the sustained increase of air traffic during the last decades has led to severe saturation and environmental problems and new needs for aircraft guidance applications have appeared: traffic avoidance and relative guidance maneuvers, formation flight, continuous descent approaches, time metering and landing maneuvers. It has become more and more difficult to integrate these new complex functions within the classical discrete mode-based approach of guidance -where complex trajectories are performed through the scheduling of elementary guidance modes. Until recently few works have been published by the flight control community to deal with what seems more profitable and safe in that case: a full trajectory-based approach of flight guidance. This communication is to propose an efficient trajectory-based guidance system with the help of flatness property of aircraft dynamics and neural networks. The proposed approach has been applied to the case of a general aviation aircraft where the inertial position of the aircraft is shown to be a differential flat output for its flight guidance dynamics. Since this differential flatness property is implicit as a result of the involved complex aerodynamic phenomena, a feed-forward neural network is built to deal with the inversion of the guidance dynamics. Once conveniently structured and trained, the neural network appears to be able to generate in real time the reference values to conventional autopilot systems concerned with attitude and engine regime. Then, in order to compensate for modeling errors and external disturbances, an additional feedback loop is introduced, where its gains can be chosen independently of the reference trajectories. Then, robustness with respect to a large range of trajectories is achieved in practical terms and a generic trajectory-based guidance system is obtained.
机译：差分平坦度是某些动态系统的特性，使得开发新工具来分析和控制复杂的非线性动态系统成为可能。某些具有平坦性的输出具有特殊的意义，因为从理论上讲，通过反转系统动力学，有可能从其计划轨迹中推导出相应的控制输入。一些论文研究了常规飞机全飞行动力学的差平度。就航空运输而言，过去几十年来空中交通的持续增长导致了严重的饱和和环境问题，并且出现了对飞机制导应用的新需求：避免交通和相对制导演习，编队飞行，连续下降进近，时间计量和着陆演习。将这些新的复杂功能集成到经典的基于离散模式的制导方法中变得越来越困难，在这种方法中，通过基本制导模式的调度来执行复杂的轨迹。直到最近，飞行控制界还很少发表有关这种情况下似乎更有利可图和更安全的著作：一种基于轨迹的飞行指导方法。该交流旨在借助飞机动力学和神经网络的平坦性，提出一种有效的基于轨迹的制导系统。所提出的方法已经应用于一般航空器的情况，在该情况下，航空器的惯性位置显示出其飞行制动力的平面差输出。由于涉及到复杂的空气动力学现象，这种不平坦的特性是隐性的，因此建立了前馈神经网络来处理制导动力学的反转。一旦进行了方便的构造和训练，神经网络似乎能够实时生成与姿态和发动机状况有关的常规自动驾驶系统的参考值。然后，为了补偿建模误差和外部干扰，引入了一个附加的反馈环路，在该环路中可以独立于参考轨迹选择其增益。然后，在实践上实现了针对大范围轨迹的鲁棒性，并且获得了基于轨迹的通用制导系统。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱：kefu@zhangqiaokeyan.com

京公网安备：11010802029741号 ICP备案号：京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有

  • 客服微信

  • 服务号