Simultaneous Propulsion System and Trajectory Optimization
Simultaneous Propulsion System and Trajectory Optimization
复制标题
同步推进系统和轨迹优化
DOI：
10.2514/6.2017-4435
复制
发表时间：
2017
期刊：
影响因子：
--
通讯作者：
J. Gray
中科院分区：
文献类型：
--
作者：
E. Hendricks;Robert Falck;J. Gray
最近提出了许多新的飞机方案，它们把推进系统的设计和操作与整个飞行器的设计和性能特性紧密地结合起来。这些概念包括推进技术，例如边界层摄入、混合动力电动推进系统、分布式推进系统和变循环发动机。对这些概念的初步研究通常采用传统的解耦方法来进行飞机设计，其中空气动力学和推进设计是先验的，表格数据用于为轨迹分析提供廉价的查找表。然而，当新的飞机方案需要考虑额外的工作参数时，如多个油门设置、迎角对推进系统的影响或推进油门设置对空气动力学的影响，产生表格数据的成本开始呈指数增长。本文提出了一种新的建模方法，无需生成表格数据，而是允许将昂贵的推进或空气动力学分析直接集成到轨迹分析模型中，从而使整个设计问题以完全耦合的方式进行优化。通过使用三种相对较新的分析工具：OpenMDAO，PyCycle和Pointer，实现一个典型的最优控制问题，F-4最小爬升时间轨迹优化，证明了新方法。PyCycle和Pointer都提供了分析导数，OpenMDAO使这两个工具能够组合成一个耦合模型，该模型可以以高效的并行方式运行，以避免更昂贵的推进分析所增加的成本。该模型产生的结果作为紧耦合设计方法的验证，并指导未来的研究，以检查飞机的概念与更复杂的操作依赖的空气动力学和推进模型。以保持壁时间的合理优化。为了使并行执行，Legendre-Gauss-Radau（LGR）的转录方法，这是服从并行化。结合J79和F-4模型，比较了一系列具有不同加力燃烧室温度极限的发动机设计的最佳轨迹。结果表明，发动机设计对飞机的爬升时间性能有很大影响，对飞机的最后弹道形状也有很大影响。总的来说，结果证明了这种新方法的可行性，它结合了LGR伪频谱轨迹分析与更昂贵的周期分析为基础的推进模型，以创建一个耦合的推进轨迹分析。结果还表明，该方法可以有效地并行执行，从而使未来的评估更先进的，耦合的航空推进系统的概念。
A number of new aircraft concepts have recently been proposed which tightly couple the propulsion system design and operation with the overall vehicle design and performance characteristics. These concepts include propulsion technologies such as boundary layer ingestion, hybrid electric propulsion systems, distributed propulsion systems and variable cycle engines. Initial studies examining these concepts have typically used a traditional decoupled approach to aircraft design where the aerodynamics and propulsion designs are done a-priori and tabular data is used to provide inexpensive look up tables to the trajectory analysis. However the cost of generating the tabular data begins to grow exponentially when newer aircraft concepts require consideration of additional operational parameters such as multiple throttle settings, angle-of-attack effects on the propulsion system, or propulsion throttle setting effects on aerodynamics. This paper proposes a new modeling approach that eliminates the need to generate tabular data, instead allowing an expensive propulsion or aerodynamic analysis to be directly integrated into the trajectory analysis model enabling the entire design problem to be optimized in a fully coupled manner. The new method is demonstrated by implementing a canonical optimal control problem, the F-4 minimum time-to-climb trajectory optimization, using three relatively new analysis tools: OpenMDAO, PyCycle and Pointer. PyCycle and Pointer both provide analytic derivatives and OpenMDAO enables the two tools to be combined into a coupled model that can be run in an efficient parallel manner to offset the increased cost of the more expensive propulsion analysis. Results generated with this model serve as a validation of the tightly coupled design method and guide future studies to examine aircraft concepts with more complex operational dependencies for the aerodynamic and propulsion models. order to keep wall times reasonable for the optimization. To enable parallel execution, a Legendre-Gauss-Radau (LGR) transcription approach which is amenable to parallelization is employed. The combined J79 and F-4 model is then used to compare optimal trajectories for a series of engine designs with different after-burner temperature limits. The results show that the engine design has a strong impact on time-to-climb performance of the aircraft and can also have a meaningful affect on the shape of the final trajectory as well. Overall, the results prove the viability of this new approach which combines LGR psue-dospectral trajectory analysis with the more expensive cycle analysis based propulsion model to create a coupled propulsion-trajectory analysis. The results also show that the method can be executed efficiently in parallel, thereby enabling future evaluation of more advanced, coupled aviation propulsion system concepts.