Dynamic Lateral Entry Guidance Logic
Dynamic Lateral Entry Guidance Logic
复制标题
DOI：
10.2514/1.8008
复制
发表时间：
2004-08
影响因子：
2.6
通讯作者：
Zuo-jun Shen;P. Lu
中科院分区：
文献类型：
--
作者：
Zuo-jun Shen;P. Lu
进入车辆的横向运动由其倾斜角度的符号控制，该倾斜角度由进入引导系统确定。用于改变银行标志的传统技术是基于车辆相对于着陆点的航向误差的预定阈值。基于横向范围开发了一种新的自动横向引导逻辑，发现横向范围是比航向误差更适合横向引导的参数。当前的引导逻辑通过不断评估来自参考横向范围轮廓、当前横向范围和估计的实际升阻比的信息来确定坡度反转。在接近轨迹末端时，非迭代数值预测器决定是否需要最终的倾斜反转来消除航向误差。该算法使进入制导系统能够执行各种任务和截然不同的倾斜角剖面，并在存在重大空气动力学建模不确定性的情况下提供可靠且良好的性能。所有这些任务都可以完成，无需手动调整引导参数或使用过多的银行反转。提出了不同任务场景和显着分散的广泛高保真模拟，以证明所提出方法的性能。 I. 简介 用于升降进入飞行器的进入引导系统在大气层飞行过程中控制倾斜角和迎角，飞行高度约为 120 公里，直到速度降至马赫 2‐3（参考文献 1）。引导命令是通过跟踪参考轨迹来生成的，该参考轨迹要么在飞行前规划中设计，1要么可能在机上设计，正如最近的努力试图实现的那样。 2−4 在传统方法中，要跟踪的参考代表所需的纵向剖面，就像航天飞机的情况一样，其中参考是阻力加速度与速度剖面，它等效地定义了距离与能量条件。 1 跟踪纵向轮廓确定倾斜角命令的大小。另一方面，每当相对于目标着陆点的航向误差超过预定阈值时，倾斜角命令的符号就会改变为相反的。这种强制改变坡度符号称为坡度角反转。当实际飞行按任务前分析中的计划进行并且实际倾斜角大小保持接近参考值时，传统方法效果很好。当飞行的坡度曲线明显不同时，坡度反转阈值标准将需要调整，通常通过模拟进行调整。可能需要非常不同的倾斜角剖面的情况包括从不同的轨道进入、在备用着陆点着陆、进入条件的变化导致不同的下行范围和横向范围以及未计划的应急进入任务（例如按需进入和中止）。轨迹规划算法的最新进展可能允许根据实际情况在机上生成参考轨迹。然而，相应的倾斜角轮廓可能与标称倾斜角轮廓有很大不同。在存在严重空气动力学模型不匹配的情况下，可能会出现类似的情况，即使是标称纵向剖面飞行也可能需要相当不同的倾斜角剖面
Lateral motion of an entry vehicle is controlled by the sign of its bank angle, determined by the entry guidance system. The conventional technique for changing the bank sign is based on prespecified threshold values in the heading error of the vehicle with respect to the landing site. A new automated lateral guidance logic is developed based on the crossrange, which is found to be a more suitable parameter than the heading error for lateral guidance. The present guidance logic determines the bank reversals by constantly evaluating information from the reference crossrange profile, current crossrange, and estimated actual lift-to-drag ratio. Near the end of the trajectory, a noniterative numerical predictor decides whether a final bank reversal is needed to null the heading error. This algorithm enables the entry guidance system to fly a wide range of missions and vastly different bank-angle profiles and provides reliable and good performance in the presence of significant aerodynamic modeling uncertainty. All of these tasks can be accomplished without requiring manual tuning of guidance parameters or using an excessive number of bank reversals. Extensive high-fidelity simulations with different mission scenarios and significant dispersions are presented to demonstrate the performance of the proposed method. I. Introduction T HE entry guidance system for a lifting entry vehicle controls the bank angle and angle of attack during the atmospheric flight from the entry interface at about altitude 120 km until the velocity decreases to Mach 2‐3 (Ref. 1). The guidance commands are generated by tracking a reference trajectory which is designed either in preflight planning, 1 or potentially on board, as recent efforts have attempted to achieve. 2−4 In the traditional approach, the reference to be tracked represents the desired longitudinal profiles as in the case of the shuttle, where the reference is a drag-acceleration-vs-velocity profile that equivalently defines the range-vs-energy condition. 1 Tracking the longitudinal profiles determines the magnitude of the bank-angle command. The sign of the bank-angle command, on the other hand, is changed to the opposite whenever the heading error with respect to the targeted landing site exceeds a prespecified threshold. This forced change of bank sign is referred to as bank-angle reversal. The traditional approach works well when the actual flight goes as planned in pre-mission analysis and the actual bank-angle magnitude remains close to the reference value. When the bank profile flown is significantly different, the bank-reversal threshold criterion will require adjustments, typically made through simulations. The situations where a very different bank-angle profile may be necessary include entry from a different orbit, landing at an alternate landing site, variations in entry conditions leading to much different downrange and crossrange, and contingency entry missions (such as on-demand entry and aborts) that are not planned. The recent advances in trajectory-planning algorithms would potentially allow on-board generation of a reference trajectory based on the actual conditions. However, the corresponding bank-angle profiles could be drastically different from the nominal one. Similar situations can develop in the presence of significant aerodynamic modeling mismatch, where flying even the nominal longitudinal profiles may necessitate a bank-angle profile considerably different