基于脉冲初值的小推力转移轨道优化研究

针对小推力转移轨道优化过程往往忽略初值多样性的现状,研究了基于不同脉冲初值的小推力转移轨道优化问题。基于直接法的离散思想建立了小推力转移轨道优化模型,提出了基于粒子群和序列二次规划的组合优化算法,以地球1∶1共振近地小行星2016HO3交会任务为例,将3种典型的脉冲轨道作为初值设计了燃料最优小推力转移轨道。仿真结果表明:3种初值轨道优化得到了2个小推力转移发射窗口,两者燃料消耗差距不超过6%。不同的初值对小推力轨道的整体性能指标影响较小,但开关机时刻和推力方向的变化会产生较大差异,从而得到不同的最优控制曲线。     

基于退火遗传算法的小推力轨道优化问题研究

利用退火遗传算法解决小推力轨道优化问题。首先利用传统混合法将轨道优化问题归结为受非线性方程约束的参数优化问题。通过结合退火和随机惩罚函数对约束条件进行处理后，用遗传算法求解这个参数优化问题。最后再采用局部优化算法提高解的精度。这种算法既保持了传统混合法精度高、解轨线光滑的优点，又克服了传统轨道优化方法收敛性差、初始猜测困难、容易陷入局部极小解的缺点。在本文的最后，利用文中提出的轨道优化算法求解“喷-停-喷”型定常推力幅值地球-木星轨道转移问题。算例证明此算法可以有效地求解小推力轨道转移问题，尤其适用于传统轨道优化方法难以求解的复杂轨道优化问题。     

Inter-orbital low-thrust transfers in an arbitrary field of forces

Low-thrust transfers between preset orbits are considered in the presence of perturbations of different origin. A simple method of finding the transfer trajectory is suggested, based on linearization of motion near reference orbits. The required accuracy of calculations is achieved by way of increasing the number of reference orbits. The method can also be used in the case of a large number of revolutions around the attracting center: no averaging of motion is required in this case. The suggested method is applicable as well, when the final orbit is specified partially and when there are constraints on the thrust direction. The optimal solution to the linearized problem is not optimal for the original problem; closeness of solutions to these two problems is estimated using a numerical example. Capabilities of the method are also illustrated by examples.     

Optimum low-thrust limited power transfers between neighbouring elliptic non-equatorial orbits in a non-central gravity field

A complete first-order analytical solution is developed for the problem of optimum low-thrust limited power transfers between neighbouring elliptic non-equatorial orbits in a non-central gravity field. The optimization problem is formulated as a Mayer problem of optimal control with Cartesian elements as state variables. After applying the Pontryagin maximum principle and determining the optimal thrust acceleration, an intrinsic canonical transformation is performed: the Cartesian elements are changed by suitable orbital elements. Hori's method is applied in determining a first-order analytical solution. Simple analytical solutions are obtained explicitly for long-time transfers.     

Optimization of low-thrust orbit transfers with constraints

Approximate numerical methods of optimization are presented for multi-orbit noncoplanar orbit transfers of low-thrust spacecraft. The linear representation of derivatives of boundary parameters is used in the vicinity of a reference trajectory with its discretization into small segments. For each segment a set of pseudo-impulses is introduced, representing possible directions of the thrust vector. In order to solve essentially nonlinear problems, the iterative process of upgrading the reference trajectory is used. At each iteration the linear programming problem of high dimensionality is solved, its boundary conditions being represented in the form of a linear matrix equation. Interval constraints are considered in the form of blocking the propulsion system operation on shadow segments of the orbit, as well as cycling constraints, and constraints on total duration of maneuvers at certain trajectory segments. The results of comparison with solutions obtained by other methods are presented together with examples illustrating the convergence of iterative processes. Optimizations of the trajectories for launching geosynchronous satellites to their orbits and of the trajectories of a noncoplanar transfer from low to high-elliptic Molniya orbit are considered under these constraints.     

Optimization of transfers under constraints on the thrust direction: II

This paper completes the study of optimal transfers with constraints imposed on the thrust vector direction that was opened by paper [1]. The linear inhomogeneous and homogeneous constraints on the thrust direction are considered (specified either by equalities or inequalities), as well as mixed constraints. Some examples of the constraints are presented. A modified method of the transporting trajectory is applied in order to find the optimal transfer under the linear constraints on the thrust direction. This method also gives the necessary condition for a transfer possibility at a given constraint on the thrust direction. A numerical example is considered, in which the propellant consumption is analyzed for the cases of transfers with and without constraints.     

Electric transfer optimization for mars sample return mission

The aim of this paper is to analyse an alternative scenario for Mars Sample Return Orbiter mission, where electric propulsion is used for Earth-Mars and Mars-Earth heliocentric cruises and for Mars orbit insertion / escape transfers, whereas chemical propulsion is used for final Mars rendezvous. The problem consists in minimizing the initial vehicle mass to obtain a specific final dry mass in reasonable time. The planetocentric phases correspond to continuous low-thrust trajectories, spiraling around Mars between a low orbit and the influence sphere altitude. The heliocentric phases consist of a succession of low-thrust and coasting arcs with specific departure and arrival conditions at the Earth. For these two types of transfer, efficient optimal control tools exist based on Pontryagin's maximum principle. Thanks to the coordination between planetocentric and heliocentric phases, the solution obtained with these two separate tools gives a good upper bound of the optimal solution in terms of propellant consumption and duration. This optimization procedure is described and finally applied to the proposed mission. The numerical results are presented and compared with the baseline chemical mission solution. The electric option could allow to decrease the spacecraft departure mass but may lead to rather long mission duration.     

Mission analysis and systems design of a near-term and far-term pole-sitter mission

This paper provides a detailed mission analysis and systems design of a near-term and far-term pole-sitter mission. The pole-sitter concept was previously introduced as a solution to the poor temporal resolution of polar observations from highly inclined, low Earth orbits and the poor high-latitude coverage from geostationary orbit. It considers a spacecraft that is continuously above either the north or south pole and, as such, can provide real-time, continuous and hemispherical coverage of the polar regions. Being on a non-Keplerian orbit, a continuous thrust is required to maintain the pole-sitter position. For this, two different propulsion strategies are proposed, which result in a near-term pole-sitter mission using solar electric propulsion (SEP) and a far-term pole-sitter mission where the SEP thruster is hybridized with a solar sail. For both propulsion strategies, minimum propellant pole-sitter orbits are designed. In order to maximize the spacecraft mass at the start of the operations phase of the mission, the transfer from Earth to the pole-sitter orbit is designed and optimized assuming either a Soyuz or an Ariane 5 launch. The maximized mass upon injection into the pole-sitter orbit is subsequently used in a detailed mass budget analysis that will allow for a trade-off between mission lifetime and payload mass capacity. Also, candidate payloads for a range of applications are investigated. Finally, transfers between north and south pole-sitter orbits are considered to overcome the limitations in observations due to the tilt of the Earth's rotational axis that causes the poles to be alternately situated in darkness. It will be shown that in some cases these transfers allow for propellant savings, enabling a further extension of the pole-sitter mission.     

小推力最优轨迹协态估计的高效机器学习方法

针对变比冲小推力轨迹间接优化中的协态变量初值猜测问题,提出了一种基于机器学习的协态变量初值高精度高效估计方法。首先,基于标称最优轨迹延拓,建立了状态量边值高扰动上限情形下的数据集生成方法,并分析了扰动上限对求解效率的影响。然后,构建了基于位置速度、轨道根数和改进春分点轨道根数多形式状态量组合输入的人工神经网络(ANN)映射关系,分析并优化了神经网络结构。将提出的方法应用于深空探测小推力转移场景,仿真结果表明该方法相对于标称轨迹直接扰动的数据集生成方法及单一形式状态量输入的人工神经网络映射方法,均有效地提升了求解收敛率,能够高效高精度地估计协态变量初值,实现轨迹快速优化。     

一种燃料最省的火星精确着陆动力下降段快速轨迹优化方法

针对火星精确着陆问题,给出一种燃料最省的火星着陆动力下降段快速轨迹优化方法。首先以燃料最省为指标建立了着陆器动力下降段轨迹优化问题模型,然后利用极大值原理对该模型的推力大小、推力方向与协状态变量之间的关系进行分析,得出着陆器推力只能以最大或最小推力工作、推力方向和与速度对应的协状态矢量方向相同的结论。并以此为基础,将原问题的状态微分方程和协状态微分方程转换为有限个推力为常值的分段函数,推导每段内微分方程的解析表达式,给出了动力下降段轨迹优化方法的算法流程;最后通过数学仿真将所提出的方法与凸规划、多项式制导方法进行比较。结果表明该方法不仅避免了多项式方法没有考虑推力约束及燃料消耗的缺陷,而且在计算效率方面较凸规划方法有大幅度提高。     

电动帆轨迹优化及其性能分析

电动帆是一种新兴的无推进剂损耗的推进方式,利用太阳风的动能冲力飞行。电动帆由数百根长而细的金属链所组成,这些金属链通过空间飞行器自旋展开,太阳能电子枪向外喷射电子,使金属链始终保持在高度的正电位,这些带电的金属链会排斥太阳风质子,利用太阳风的动能冲力推动空间飞行器驶向目标方向。针对电动帆轨迹优化问题,提出采用Gauss伪谱法进行轨迹优化,克服了间接法对协态变量初值敏感的缺点。考虑在太阳风暴等原因造成特征加速度改变的情况,基于Gauss伪谱法实现电动帆在线轨迹重新规划,提高电动帆对太阳风不确定性的适应能力。最后以太阳系外探测任务为例,对电动帆和太阳帆的性能进行对比,仿真结果表明电动帆在星际远航任务中所用时间较短。     

High-thrust and low-thrust two-stage LEO-LEO transfer

In this paper a low-altitude orbit-to-orbit minimum-fuel transfer is discussed. The spacecraft consists of a high-thrust solid stage and a low-thrust liquid stage. The thrust acceleration ratio is greater than 500. Both initial and final orbits are circular but non-coplanar. In particular, altitudes of 300 and 500–600 km together with an inclination difference of about 16 deg are considered. J2 and drag perturbations and flight constraints are taken into account. The current discussion is centred on the nominal trajectory of a case of real interest.     

基于轨迹成型法星际小推力转移轨道快速设计

给出了一种基于轨迹成型法星际小推力转移轨道的快速设计方法.首先介绍了基于轨迹成型法的有关内容,并通过与Hohmann变轨相对比验证了该方法描述小推力变轨的可用性.鉴于这种方法数学模型收敛性强,计算速度快的特点,以地火转移轨道为例,提出了不同推力条件下发射窗口的搜索方法.最后,利用基于轨迹成型法和真实数值仿真之间误差小、线性度好的特点,通过几次简单的线性迭代设计出小推力转移轨道.     

A low-thrust transfer between the Earth–Moon and Sun–Earth systems based on invariant manifolds

A low-energy, low-thrust transfer between two halo orbits associated with two coupled three-body systems is studied in this paper. The transfer is composed of a ballistic departure, a ballistic insertion and a powered phase using low-thrust propulsion to connect these two trajectories. The ballistic departure and insertion are computed by constructing the unstable and stable invariant manifolds of the corresponding halo orbits, and a complete low-energy transfer based on the patched invariant manifolds is optimized using the particle swarm optimization (PSO) algorithm on the criterion of smallest velocity discontinuity and limited position discontinuity (less than 1 km). Then, the result is expropriated as the boundary conditions for the subsequent low-thrust trajectory design. The fuel-optimal problem is formulated using the calculus of variations and Pontryagin's Maximum Principle in a complete four-body dynamical environment. Then, a typical bang–bang control is derived and solved using the indirect method combined with a homotopic technique. The contributions of the present work mainly consist of two points. Firstly, the global search method proposed in this paper is simply handled using the PSO algorithm, a number of feasible solutions in a fairly wide range can be delivered without a priori or perfect knowledge of the transfers. Secondly, the indirect optimization method is used in the low-thrust trajectory design and the derivations of the first-order necessary conditions are simplified with a modified controlled, restricted four-body model.     

极坐标系连续常值推力机动分析

连续常值推力是空间飞行常用的轨道机动方式，在空间交会与星际航行使命中具有重要的应用价值。其中，小推力适合于地球轨道航天器交会机动，而切向或周向推力以及较大的正径向推力可用于脱离地球引力场的逃逸飞行，执行星际交会使命。应用常推力作用下的极坐标系质心运动方程，对机动推力的量值没有限制；在航天器交会应用中，对相对距离也无要求。这种方法可直接获得向径与速度等轨道参数随时间或极角（绕地心的转动角）的变化，便于分析轨道转移与逃逸运动，有助于飞行使命与运动轨迹的设计。特别是，若机动转移的初轨为圆轨道，在推力较小、飞行时间不长的情况下，应用无量纲形式运动方程，可获得具有工程应用价值的近似解。文章给出一些有关的结果与应用案例。     

Design of optimal earth pole-sitter transfers using low-thrust propulsion

Recent studies have shown the feasibility of an Earth pole-sitter mission using low-thrust propulsion. This mission concept involves a spacecraft following the Earth's polar axis to have a continuous, hemispherical view of one of the Earth's poles. Such a view will enhance future Earth observation and telecommunications for high latitude and polar regions. To assess the accessibility of the pole-sitter orbit, this paper investigates optimum Earth pole-sitter transfers employing low-thrust propulsion. A launch from low Earth orbit (LEO) by a Soyuz Fregat upper stage is assumed after which solar electric propulsion is used to transfer the spacecraft to the pole-sitter orbit. The objective is to minimize the mass in LEO for a given spacecraft mass to be inserted into the pole-sitter orbit. The results are compared with a ballistic transfer that exploits manifold-like trajectories that wind onto the pole-sitter orbit. It is shown that, with respect to the ballistic case, low-thrust propulsion can achieve significant mass savings in excess of 200 kg for a pole-sitter spacecraft of 1000 kg upon insertion. To finally obtain a full low-thrust transfer from LEO up to the pole-sitter orbit, the Fregat launch is replaced by a low-thrust, minimum time spiral, which provides further mass savings, but at the cost of an increased time of flight.     

Trajectories for missions of low-thrust spacecraft aimed at delivery of soil samples from main belt asteroids and Phobos

Trajectories of spacecraft with electro-jet low-thrust engines are studied for missions planning to deliver samples of matter from small bodies of the Solar System: asteroids Vesta and Fortuna, and Martian moon Phobos. Flight trajectories are analyzed for the mission to Phobos, the limits of optimization of payload spacecraft mass delivered to it are determined, and an estimate is given to losses in the payload mass when a low-thrust engine with constant outflow velocity is used. The model of an engine with ideally regulated low thrust is demonstrated to be convenient for calculations and analysis of flight trajectories of a low-thrust spacecraft.     

基于傅立叶平均法下的连续小推力动力学分析

通过将小推力展开为偏近点角的傅立叶级数,并对高斯摄动方程在一个轨道周期上的平均,将原方程的推力转化为仅由14个傅立叶系数表示的控制变量。仿真计算表明,平均化后的高斯方程使计算量与牛顿积分相比显著减少,且对小推力而言有足够的精度。对利用平均化后的高斯方程计算轨道根数时产生误差的原因进行了研究,并进一步分析小推力的范围和小推力近似表达式对上述误差的影响,为今后小推力下非开普勒轨道动力学分析提供了理论依据和参数。     

基于双蚁群优化的快速轨道转移

研究了Lambert转移的优化问题,并用于航天器的快速轨道机动。以二体模型为基础,在给定的转移时间范围内以燃料最省为目标函数,用蚁群算法(ACA)对Lambert机动问题进行寻优;将第一次寻优用于完成轨道机动所需的速度增量转换成航天器发动机有限推力下的时间序列;考虑地球主要摄动影响,再次用ACA对有限推力工作时段进行边值修正,保证航天器能转移至目标位置范围。仿真结果表明:双蚁群优化方法可有效修正由二体假设导致的Lambert转移的误差,同时由蚁群算法的较强寻优性能保证了Lambert机动过程的燃料最省。