基于遗传算法的RLV再入轨迹优化设计

飞行器再入轨迹优化是一类最优控制问题。传统的优化方法存在初始值敏感问题,利用改进的基于小生境技术和精英方法的适应值共享拥挤遗传算法进行RLV的再入轨迹优化设计。以终端时间固定的空间最小控制能量再入轨迹和终端时间自由的平面最小热载再入轨迹为例,详细讨论了遗传算法用于再入轨迹优化设计所需要解决的一些关键问题。仿真结果表明提出的方法能够较快地搜索到全局最优解,对初始猜测值不敏感,能够方便用于RLV的再入轨迹方案选择和优化设计。     

亚轨道飞行器返回轨迹快速优化

快速、准确、鲁棒的轨迹生成方法可以增加任务的安全性与可靠性,极大地降低成本。针对亚轨道飞行器返回段特点,引入"伪控制量"、"末端进场走廊"等概念,分别采用高斯伪谱法和向前拉道伪谱法进行了返回轨迹快速优化研究,比较了两种伪谱法在处理复杂问题时的能力。仿真结果表明,向前拉道伪谱法不适合处理含控制量约束的问题,而高斯伪谱法在满足各种约束条件下,能够快速准确地生成亚轨道飞行器返回轨迹,同时验证了结果的可行性与最优性。     

Conditions of the Maximum Principle in the Problem of Optimal Control over an Aggregate of Dynamic Systems and Their Application to Solution of the Problems of Optimal Control of Spacecraft Motion

The necessary first-order conditions of strong local optimality (conditions of maximum principle) are considered for the problems of optimal control over a set of dynamic systems. To derive them a method is suggested based on the Lagrange principle of removing constraints in the problems on a conditional extremum in a functional space. An algorithm of conversion from the problem of optimal control of an aggregate of dynamic systems to a multipoint boundary value problem is suggested for a set of systems of ordinary differential equations with the complete set of conditions necessary for its solution. An example of application of the methods and algorithm proposed is considered: the solution of the problem of constructing the trajectories of a spacecraft flight at a constant altitude above a preset area (or above a preset point) of a planet's surface in a vacuum (for a planet with atmosphere beyond the atmosphere). The spacecraft is launched from a certain circular orbit of a planet's satellite. This orbit is to be determined (optimized). Then the satellite is injected to the desired trajectory segment (or desired point) of a flyby above the planet's surface at a specified altitude. After the flyby the satellite is returned to the initial circular orbit. A method is proposed of correct accounting for constraints imposed on overload (mixed restrictions of inequality type) and on the distance from the planet center: extended (nonpointlike) intermediate (phase) restrictions of the equality type.     

Guidance of a spacecraft with low lift-to-drag ratio to a landing point

The problem of terminal control over a deorbiting spacecraft at the stage of its flight after leaving plasma (altitude of ∼40 km) is considered, the aim being to guide it to a preset landing point. The algorithm is based on a modification of the well-known method of proportional navigation, when a fixed point is the target. It is suggested to use satellite navigation systems (of the GLONASS or GPS types) and/or radio beacons, which should allow one to determine the spacecraft trajectory parameters with high precision. Single-channel control is performed by changing the roll angle according to current parameters of the trajectory, which ensures adaptability of the method. Examples of three-dimensional trajectories of flight are presented for a manned spacecraft with low lift-to-drag ratio (∼0.5), currently under design in Russia. The results of statistical modeling taking into account initial deviations of the trajectory parameters and wind disturbances are presented. A method of statistical choice of a reference trajectory for the guidance stage is suggested. A theoretical possibility of using the algorithm of spacecraft guidance (in case of in-light accident with a carrier launcher) to preset regions in the vicinity of launching route is demonstrated. A qualitative analysis of proportional navigation with a fixed target is presented.     

三维弹道有理Bezier曲线造型与优化方法

为有效克服传统弹道优化方法解的收敛性和最优性受搜索算法、初始猜测等影响大的问题,提出了基于有理Bezier曲线的三维弹道造型与优化计算方法。根据边界条件用光滑且只含少量造型变量的有理Bezier曲线形成待优化三维造型弹道,采用逆动力学方法计算造型弹道倾角、弹道角速度和法向加速度等,结合弹道其它动力学方程和约束计算造型弹道性能指标,通过对弹道造型变量寻优得到最优弹道。这种方法将无限维弹道优化问题转换为对很少造型变量的参数优化问题, 而且得到的最优弹道完全光滑可飞。与间接法相比,无需求解两点边值问题使用方便、鲁棒性强;与直接法相比,不需要对动力学方程离散化、寻优变量少、求解收敛性好,而且解的全局最优性和光滑性更高。仿真结果及与自适应伪谱法的比较校验了方法的实用性和有效性。     

Approximate solutions to minimax optimal control problems for aeroassisted orbital transfer

This paper considers minimax problems of optimal control arising in the study of aeroassisted orbital transfer. The maneuver considered involves the coplanar transfer from a high planetary orbit to a low planetary orbit. An example is the HEO-to-LEO transfer of a spacecraft, where HEO denotes high Earth orbit and LEO denotes low Earth orbit. In particular, HEO can be GEO, a geosynchronous Earth orbit.The basic idea is to employ the hybrid combination of propulsive maneuvers in space and aerodynamic maneuvers in the sensible atmosphere. Hence, this type of flight is also called synergetic space flight. With reference to the atmospheric part of the maneuver, trajectory control is achieved by means of lift modulation. The presence of upper and lower bounds on the lift coefficient is considered.The following minimax problems of optimal control are investigated: (i) minimize the peak heating rate, problem P1; and (ii) minimize the peak dynamic pressure, problem P2. It is shown that problems P1 and P2 are approximately equivalent to the following minimax problem of optimal control: (iii) minimize the peak altitude drop occurring in the atmospheric portion of the trajectory, problem P3.Problems P1–P3 are Chebyshev problems of optimal control, which can be converted into Bolza problems by suitable transformations. However, the need for these transformations can be bypassed if one reformulates problem P3 as a two-subarc problem of optimal control, in which the first subarc connects the initial point and the point where the path inclination is zero, and the second subarc connects the point where the path inclination is zero and the final point: (iv) minimize the altitude drop achieved at the point of junction between the first subarc and the second subarc, problem P4. Note that problem P4 is a Bolza problem of optimal control.Numerical solutions for problems P1–P4 are obtained by means of the sequential gradient-restoration algorithm for optimal control problems. Numerical examples are presented, and their engineering implications are discussed. In particular, it is shown that, from an engineering point of view, it is desirable to solve problem P3 or P4, rather than problems P1 and P2.     

Definition and analysis of properties of optimal three-impulse point-to-orbit transfers under time constraint

This paper deals with energetically optimal multi-impulse transfers of a spacecraft in the central Newtonian gravitational field near a planet. The transfer from a point on initial orbit to the final orbit with the given angular momentum and energy constants is considered. The transfer time is bounded above.With the distance from spacecraft to planet limited and the time free, such parameters of given orbits are chosen that the 3-impulse apsidal transfer T_r is optimal with an intermediate impulse at the maximum distance. On the basis of necessary optimality conditions an algorithm is developed to numerically determine the desired optimal transfer trajectory T_t under time constraint, the apsidal trajectory T_r being taken as initial approach. From the geometry and energy viewpoints, both trajectories T_t and T_r are close to each other. The trajectory T_t is also 3-impulsive, all impulses on it are nonapsidal. The distance from the planet is larger and the sum of impulses is less for this trajectory than for the initial trajectory T_r with the same transfer time.The simplified solution of the problem is constructed producing good approximation to the exact numerical optimization results. The solution asymptotics is found when the transfer time tends to infinity.     

Optimal ballistically captured Earth–Moon transfers

The optimality of a low-energy Earth–Moon transfer terminating in ballistic capture is examined for the first time using primer vector theory. An optimal control problem is formed with the following free variables: the location, time, and magnitude of the transfer insertion burn, and the transfer time. A constraint is placed on the initial state of the spacecraft to bind it to a given initial orbit around a first body, and on the final state of the spacecraft to limit its Keplerian energy with respect to a second body. Optimal transfers in the system are shown to meet certain conditions placed on the primer vector and its time derivative. A two point boundary value problem containing these necessary conditions is created for use in targeting optimal transfers. The two point boundary value problem is then applied to the ballistic lunar capture problem, and an optimal trajectory is shown. Additionally, the problem is then modified to fix the time of transfer, allowing for optimal multi-impulse transfers. The tradeoff between transfer time and fuel cost is shown for Earth–Moon ballistic lunar capture transfers.     

能量最优的航天器连续动态避障轨迹规划

针对传统脉冲避障算法在航天器轨迹规划应用中存在对瞬时推力依赖性强且燃料消耗量大的问题,提出能量最优的连续动态避障算法。该算法首先基于线性相对运动方程与有限时间的能量最优模型,建立了相对运动能量最优模型,同时验证了模型最优性;其次将动态障碍物的 y 向运动误差偏移与正态分布概率引入避碰安全距离模型,修正了追踪航天器动态避障的范围,确定了安全距离矢量长度,增强了规避障碍的可靠性;最后通过障碍物速度矢量与追踪器航天器速度矢量夹角确定动态避障点方向,减少燃料消耗的同时提高了避障的有效性、准确性。通过仿真验证,该算法可以自适应选取规避障碍点,有效规避动态障碍;工质燃料消耗较小,有效延长航天器在轨寿命。     

分析插值误差和斜率的轨迹优化网格细化方法

提出一种基于插值误差和斜率分析的轨迹优化自适应网格细化方法，包括节点插入算法和节点删除算法。节点插入算法分析各个离散节点的控制变量的插值误差。若插值误差较大，则在该节点周围增加节点细化网格；否则，不进行细化。节点删除算法分析各个离散节点处的控制变量斜率。若某个节点的左斜率和右斜率都为零，那么删除该节点；否则，保留该节点。采用三个典型的轨迹优化算例验证了所提出的方法的有效性和特色，并且与其它几种网格细化方法进行了对比。仿真结果表明，本文方法生成的网格规模较小，需要的网格迭代次数较少，能够快速、高精度求解非光滑轨迹优化问题。     

基于节点改善策略的伪谱轨迹优化

基于Legendre Gauss Lobatto（LGL）伪谱法(PSM)求解最优控制问题的原理,研究了有限推力轨迹快速优化设计以及计算节点和效率改善策略。对远程变轨动力学模型进行了无量纲化处理,并设计了初值生成和串行优化求解策略,用来提高多变量多约束大规模稀疏非线性规划问题的收敛性。结合极小值原理推导了问题的最优性必要条件,基于协态映射原理设计了数值解的最优性验证方法。利用二阶微分方程技术消去部分状态量,建立了节点和计算效率改进模型,并讨论了该策略的相关数值处理技术。仿真结果表明,本文的算法可以快速地提供应用于实际飞行任务的最优解,节点改善策略在保证计算精度的同时,有效降低了计算收敛时间。     

探月返回飞行器跳跃式再入轨迹优化

基于节点自适应稀疏配点法,提出一种高精度求解探月返回飞行器跳跃式再入轨迹优化问题的方法.该方法的基本策略是:首先,应用节点自适应稀疏配点法对完整的跳跃式再入轨迹进行优化;然后,根据优化得到的控制变量对再入动力学方程进行数值积分;当积分至跳跃轨迹的最高点时,以积分得到的状态变量值作为新的初始条件,对二次再入轨迹重新优化....     

多约束条件下高超声速滑翔飞行器轨迹优化

研究了考虑航路点、禁飞区、热流、动压、过载、控制量以及终端状态等多种约束条件下高超声速滑翔飞行器轨迹优化设计问题。分析了采用一般Gauss伪谱方法进行高超声速滑翔飞行器轨迹优化设计存在的主要问题,为此,提出了轨迹分段优化策略,将轨迹优化的一般最优控制问题转换为多段最优控制问题,进而将各段轨迹按Gauss伪谱方法进行离散化,将连续多段最优控制问题转换为非线性规划问题进行求解。以多约束条件下最大射程轨迹优化为例进行仿真分析,结果表明分段优化方法能够较快地设计出满足各种约束条件的优化轨迹。     

飞行器低空突防中的威胁航线优化技术研究

低空突防中的威胁回避技术是TF／TA^2系统的核心，直接关系到低空突防飞行器的任务生存能力。针对威胁回避的特殊要求，本文提出了一种基于地形可视性分析的威胁危险指标快速计算方法，解决了威胁的量化问题。针对在大范围任务区域内进行威胁航线优化存在计算复杂性和收敛性等问题，本文尝试采用遗传算法对威胁进行优化处理。该方法得到的威胁航线，严格经过出发点和目标点，且尽量远离威胁，有效提高了飞行器低空突防的任务生存率。     

Optimization of passive orbit with the use of gravity maneuver

A high-precision method of calculating gravitational interactions is applied in order to determine optimal trajectories. A number of problems, necessary for determination of optimal parameters at a launch of a spacecraft and during its flyby near celestial bodies, are considered. The spacecraft trajectory was determined by numerical integration of the equations of passive motion of the spacecraft and of the equations of motion for planets, the Sun, and the Moon. The optimal trajectory of the spacecraft approaching the Sun is determined by fitting its initial conditions.     

全太阳系深空探测轨迹全局优化(英文)

针对太阳系中全部的248997颗行星的探测问题,给出了一种关于探测飞行器的深空探测全局四维轨迹(t,x,y,z)优化方案,即飞行器从地球发射进入太阳系并采用小推力控制,优化方案的性能指标为飞行器与太阳系中全部行星中相遇和交会的星的数量最多并且燃料消耗最少。本方案给出了四维飞行轨迹进行全局优化的一套算法,该算法由搜索算法和四维轨迹优化算法组成。此搜索算法从太阳系的248997颗行星中寻找获得尽可能多的经过近地球3维走廊内的行星;而四维轨迹优化算法由改进的动态规划算法、基于最优控制理论的共轭梯度算法和静态参数优化算法组成,其中静态参数优化算法用于搜索最优发射时间窗口。基于该组合算法,通过长时间的大规模的飞行数字仿真,最终计算出探测器的四维最优飞行轨迹,在一年内路过了太阳系中全部行星中的12颗行星。     

适应于RBCC运载器的轨迹优化建模研究

根据RBCC动力系统工作原理,分析了RBCC运载器上升段轨迹特点和设计矛盾。针对此类飞行器工作模态多、冲压模态约束条件复杂、传统优化模型无法求解其全局最优解的问题,以Radau伪谱法为基础,建立了基于混合积分变量的全程最优轨迹求解模型,并引入连接条件模型。该模型可以有效处理RBCC运载器多段飞行轨迹约束条件,克服了传统建模方法的不足。采用50km、Ma8运载任务算例校验优化模型,优化结果显示,所研究模型符合RBCC运载器工作特点和任务需要,可快速求解此类运载器上升段最优轨迹。     

Quaternion regularization in celestial mechanics and astrodynamics and trajectory motion control. I

Regularization problems in celestial mechanics and astrodynamics are considered. The fundamental regular quaternion models of celestial mechanics and astrodynamics are presented. It is shown that the efficiency of analytical investigation and numerical solution of boundary problems of optimal trajectory motion control of spacecraft may be increased using quaternion astrodynamics models. The regularization problem of celestial mechanics and astrodynamics that implies eliminating the feature, which arises in the equations of the two-body problem in case of impact of the second body with the central body, is considered in the first section of the paper. The quaternion method for regularizing the equations of the perturbed spatial two-body problem suggested by the author is presented; the method is compared with Kustaanheimo-Stiefel (KS) regularization. Demonstrative geometric and kinematic interpretations of regularizing transformations are provided. Regular quaternion equations for the two-body problem, which generalize the regular Kustaanheimo-Stiefel equations, as well as regular equations in quaternion osculating elements and quaternion regular equations for perturbed central motion of a material point, are considered. The papers on quaternion regularization in celestial mechanics and astrodynamics are briefly analyzed.     

吸气式高速飞行器爬升-巡航轨迹在线优化

针对吸气式高速飞行器的大气层内爬升和巡航飞行段,进行了在线轨迹优化设计。与传统的爬升段结合定高定速巡航轨迹形式不同,采用优化的方法得到的轨迹能够保证性能指标的最优性。首先,将轨迹优化问题建模为非线性最优控制问题,控制量包括攻角、倾侧角以及燃油当量比。然后,对问题的非线性进行了处理,将变量进行离散化,得到凸优化问题。分析了飞行器的气动和推力特性,设计了优化变量的初始参考。最后,设计算法的外环迭代策略,将每一次求得的解作为参考轨迹进行下一次迭代计算。数值仿真结果表明,该方法能够解决吸气式高速飞行器爬升巡航段轨迹优化问题,并且求解时间满足在线轨迹优化要求。