Bewegungsmodell-basierte Filtertechniken zur Navigation mit miniaturisierten Sensoren

基于运动模型的微型传感器导航过滤技术

• 批准号: 175142409
• 负责人: Professor Dr.-Ing. Florian Holzapfel
• 金额: --
• 依托单位: Institut für Raumfahrtsysteme
• 依托单位国家: 德国
• 项目类别: Research Grants
• 财政年份: 2010
• 资助国家: 德国
• 起止时间: 2009-12-31 至 2013-12-31
• 项目状态: 已结题
• 来源: https://gepris.dfg.de/gepris/projekt/175142409?language=en
• 关键词: Bewegungsmodell; basierte; Filtertechniken; zur; Navigation

In the first part of the project, the potential of model-aided navigation for dealing with longer time windows of GPS-unavailability was clearly demonstrated. Based on the development of the efficient Unified Model method by the institutes ITE and FSD and the precise understanding of model-aided navigation acquired so far, the following are identified to be the most important points to be addressed in further cooperative research:What are the commonalities and differences of the Unified Model method and the methods presented in literature?Various methods of model-aided navigation have been presented in literature. It is expected that a detailed analysis that compares the methods and identifies common and individual characteristics will lead to a deeper understanding and provide a starting point for further improvement.How to design a model-aided integrated navigation system?Because the method of model-aided navigation by itself is only a part of the navigation system, the design steps that deal with requirements of system performance (precision, robustness, cost) and restrictions (available model quality, sensors, computer power) should be studied.What extensions to the method of model-aided navigation can improve its performance?The performance of model-aided navigation suffers from inevitable errors in available models and wind estimation. These limitations should be addressed by modified or extended motion- and error-models in order to reduce the impact of modeling uncertainty. Furthermore, the model-aided navigation system should be made robust in order to deal with GPS spoofing and damages to the aircraft that reduce the validity of the used model.What is the best potential use of model-aided navigation?Depending on the used platform, the available model and IMU quality, the application scenario (length, GPS availability ...), and the desired system output it should be studied if model-aided navigation can fill a gap that has not been covered by existing standard navigation technology. The goal is to show if there is a way in which model-aided navigation could become a standard technology in UAVs or small general aviation aircraft.

在该项目的第一部分，清楚地展示了模型辅助导航在处理GPS不可用的较长时间窗口方面的潜力。基于ITE和FSD对有效的统一模型方法的发展和到目前为止对模型辅助导航的准确理解，确定了以下几点是进一步合作研究中需要解决的最重要的问题：统一模型方法与文献中提出的方法有什么共同点和不同点？文献中已经介绍了各种模型辅助导航方法。如何设计模型辅助组合导航系统？由于模型辅助导航方法本身只是导航系统的一部分，因此需要研究满足系统性能要求(精度、稳健性、成本)和约束条件(可用的模型质量、传感器、计算机能力)的设计步骤。模型辅助导航方法的哪些扩展可以提高其性能？模型辅助导航方法的性能受到现有模型和风估计的不可避免的误差影响。这些限制应该通过修改或扩展运动模型和误差模型来解决，以减少建模不确定性的影响。此外，模型辅助导航系统应该变得健壮，以应对GPS欺骗和对飞机的损坏，从而降低所用模型的有效性。模型辅助导航的最佳潜在用途是什么？取决于使用的平台、可用的模型和IMU质量、应用场景(长度、GPS可用性...)和期望的系统输出。如果模型辅助导航可以填补现有标准导航技术所未涵盖的空白，则应进行研究。其目标是展示是否有一种方法可以使模型辅助导航成为无人机或小型通用航空飞机的标准技术。