约束处理策略对遗传算法优化性能的影响

根据遗传算法处理带约束优化问题时常用的拒绝策略或惩罚策略的优势与不足,提出一种“自适应拒绝＋惩罚”相结合的约束处理策略,分别采用3种不同约束处理策略完成了空中发射运载火箭的优化设计,并对优化结果进行了比较.结果表明,对于约束严的优化问题,拒绝策略收敛速度较慢;惩罚策略中罚函数的形式对优化结果影响较大;“自适应拒绝＋惩罚”策略拒绝所有约束破坏程度超过容许范围的非可行解,并对约束破坏程度在容许范围内的非可行解进行惩罚,可大大改善遗传算法处理严约束优化问题的优化性能.     

一种卫星成像调度的约束修正方法

约束修正是卫星成像调度的重要组成部分,负责处理成像调度方案的约束满足与优化的问题.为得到优化可行的成像调度方案,针对一类带有卫星存储器和数据下传等约束的卫星成像调度问题,提出一种新的约束修正方法.首先构建时间序有向图模型,将约束修正问题归结为点带约束成本的路径搜索问题,针对该模型提出一种基于标记更新的约束修正算法.通过针对实际问题的实验分析,表明该方法能够在规定的时间内求得问题的最优解,有效的解决了卫星成像调度的约束满足与优化问题.     

有限推力交会的最省燃料轨迹

给出了航天器有限推力交会的最省燃料轨迹。首先应用三角变换技术将推力约束 转化为没有任何约束的虚拟控制,进而利用直接优化方法,应用参数化控制方法以及强化技 术将控制向量表示为分段常值函数,将上述最优控制问题转化为非线性规划问题。应用经典 的参数优化方法即可求得最优控制律的一个近似解,通过增加参数个数,重复优化得到逼近 连续最优解的参数化解。仿真结果表明提出的控制方案是行之有效的。
     

刚体航天器的最小能量姿态机动最优控制研究

针对航天器姿态机动最优控制问题.研究了初始条件和目标状态给定,以三个正交安装的动量轮为控制执行机构且机动控制时间不定的最小能量姿态机动控制方案.首先,采用与传统打靶法不同的非线性规划NLP方法来描述带约束的姿态控制问题,将控制次数设为固定值,将采样周期看作变量,通过迭代方法进行求解;然后,提出NLP初始可行解的构造方法,以不同初始可行解为初值,多次寻优,最终找到刚体航天器最小能量姿态机动的相对最优控制方案;最后,通过仿真试验验证了该方法的可行性和优越性.     

月球精确定点软着陆轨道设计及初始点选取

研究了一种应用参数化控制求解月球探测器精确定点软着陆最优控制问题的方法。 首 先用约束变换技术将不等式约束进行了近似处理,而后利用若干个分段的常数去逼近最优解 ,再根据强化技术通过时间轴上的变换,将每一段参数的持续时间转变为一组新的参数,于 是最优控制问题被转化为一系列参数优化问题。最后应用经典的参数优化方法即可求得最优 控制函数的一个近似解,通过增加参数个数,重复优化得到逼近连续最优解的参数化解。同 时在优化过程中考虑了制动初始点的选取对结果的影响。仿真结果表明了所提设计方法是简 单、有效的。〖JP〗     

基于Gauss伪谱法的固体运载火箭上升段轨迹快速优化研究

研究Gauss伪谱法在多级固体运载火箭上升段轨迹快速优化设计中的应用。针对多
级固体运载火箭上升段轨迹优化的特点,研究了求解多阶段最优控制问题的Gauss伪谱法,
引入连接点概念处理间断点,设计了边界控制变量计算方法。为进一步提高轨迹优化速度,
设计了含初值生成器的Gauss伪谱法串行轨迹优化策略,实现了固体运载火箭上升段轨迹快
速优化。仿真结果表明,采用提出的优化方法优化一条上升段轨迹所用时间为2~3分钟,终
端约束和路径约束均得到很好满足,算法求解精度高,对初值依赖性小,设计的边界控制变
量计算方法可行。     

改进差分进化算法求解多成像卫星调度问题

针对多成像卫星联合调度规划建模难度大和求解复杂度高等问题,通过分析成像卫星的成像过程和工作原理,将成像卫星调度过程分为调度预处理、任务规划和调度优化3个阶段。在调度规划过程中,建立了多星联合调度约束满足最优化模型,采用启发式算法思想,定义了个体适应度评估函数,设计了任务冲突消解方法,提出了一种改进的差分进化算法。在此基础上,采用一些确定性规则对调度规划方案可行解进行了评估和二次优化。结果表明:提出的成像卫星调度问题求解方法能够有效地分配卫星资源,生成优化的调度方案。设计结果也能够为卫星系统最优化设计和效能评估提供必要的决策支持。     

多级备件供需状态下装备使用可用度的优化

装备通常采用三级保障体制,具有多种备件供需状态,采用马尔可夫过程建立装备使用可用度(AO)模型是一个三维具有不等式约束的非线性方程。依此模型难于直观地选定AO的设计参数。本文采用优化设计方法之一的复形法较好地解决了这个问题,在λ,μ,б确定的可行域中,获得了AO的优化解及相应的λ,μ,б数值。并用实例说明优化方法可方便的解决装备论证中参数的选取问题。     

一种新的多学科设计优化方法

针对协同优化方法自身结构存在的缺陷,对其加以改进,提出一种新的单级多学科设计优化方法(ACO)。将协同优化方法中的系统相容性约束作为另一个目标函数与原来系统级目标函数交替进行优化,从而直接避免了原协同优化方法中由于相容性等式约束而带来的系统级优化求解计算困难等问题。测试结果表明新方法是可行的。     

火箭垂直回收多阶段最优轨迹规划方法

针对火箭高空再入定点回收，基于凸优化方法提出一种考虑气动力和推力控制的多阶段轨迹优化方法。在气动减速段，通过控制总攻角，实现气动升力和阻力的调制。由于气动力连续变化，使用Legendre-Gauss-Radau伪谱离散方法进行离散化，利用较少的离散点实现较高的数值精度。在动力减速段，推力矢量为控制变量。由于推力调节可能出现不连续，采用等距离散方法进行离散。在此基础上，将发动机开、关机时间也作为优化变量，并考虑各种约束，构建了多阶段离散最优控制模型。使用无损凸化方法对升力约束和推力约束进行松弛，并通过逐次凸化消除由气动力、自由时间变量以及质量引入的非凸约束，最终将问题描述为序列迭代求解的二阶锥规划问题(SOCP)。通过仿真校验，经过少量的逐次凸化迭代，可快速收敛到最优解，且落点调节范围更大，燃料更省。     

智能优化算法及其在飞行器优化设计领域的应用综述

在对国内外相关文献进行系统研究的 基础上,阐述了飞行器优化设计领域应用较为广泛的几种智能优化算法的基本原理,分析了 算法的优缺点和改进方法,总结了算法在飞行器轨迹、气动和控制等学科优化设计中的应用 情况。同时,文章对在飞行器优化设计领域应用刚刚起步,但较有发展前景的智能优化算法的 特点和应用情况进行了概述。最后,对智能优化算法在飞行器优化设计领域应用的未来研究 方向进行了分析。
     

Simple method for performance evaluation of multistage rockets

Multistage rockets are commonly employed to place spacecraft and satellites in their operational orbits. Performance evaluation of multistage rockets is aimed at defining the maximum payload mass at orbit injection, for specified structural, propulsive, and aerodynamic data of the launch vehicle. This work proposes a simple method for a fast performance evaluation of multistage rockets. The technique at hand is based on three steps: (i) the flight-path angle at each stage separation is guessed, (ii) the spacecraft velocity is maximized at the first and second stage separation, and (iii) for the last stage the thrust direction is obtained through the particle swarm optimization technique, in conjunction with the use of the Euler–Lagrange equations and the Pontryagin minimum principle. The coast duration at the second stage separation is optimized as well. The method at hand is extremely simple and easy-to-implement, but nevertheless it proves to be capable of yielding near-optimal ascending trajectories for a multistage launch vehicle with realistic structural, propulsive, and aerodynamic characteristics. The solutions found with the technique under consideration can be employed either for a rapid evaluation of the multistage rocket performance or as guesses for more refined optimization algorithms.     

基于混合法和多打靶法的有限推力航天器协同交会优化(英文)

有限推力航天器的协同交会问题是一个高维非线性的最优控制问题,传统求解方法难以收敛到最优解或者时间花费巨大。采用混合法和多打靶法构造了有限推力航天器双主动交会的数学模型,讨论了其实现最优控制的必要条件,求解了反平方力场中的最优控制数值解。推进剂总消耗最少和有限推进剂约束下的最短时间交会2种不同的算例表明,这种方法可极大地提高收敛性,快速有效地求解有限推力航天器的协同交会问题。     

Time-optimal maneuvering control of a rigid spacecraft

The time-optimal rest-to-rest maneuvering control problem of a rigid spacecraft is studied in this paper. By utilizing an iterative procedure, this problem is formulated and solved as a constrained nonlinear programming (NLP) one. In this novel method, the count of control steps is fixed initially and the sampling period is treated as a variable in the optimization process. The optimization object is to minimize the sampling period below a specific minimum value, which is set in advance considering the accuracy of discretization. To generate initial feasible solutions of the NLP problem, a genetic-algorithm-based is also proposed such that the optimization process can be started from many different points to find the globally optimal solution. With the proposed method, one can find a time-optimal rest-to-rest maneuver of the rigid spacecraft between two attitudes. To show the feasibility of the proposed method, simulation results are included for illustration.     

基于循环平稳约束的盲信号提取算法

 特定信号提取是通信信号处理中的重要问题。在约束独立分量分析方法的基础上提出了基于循环平稳约束的盲信号提取算法。该算法利用通信信号固有的循环平稳特性作为约束条件,结合独立分量分析的学习过程,同时完成特定目标信号的分离和选取,避免了经典独立分量分析方法提取“无关”源信号的过程。仿真结果表明,相比传统ICA算法,本文算法能有效提取目标源信号且提高了目标源信号的估计精度。     

基于泛化熵随机森林的雷达目标航迹起始方法

文章提出了一种基于泛化熵随机森林的雷达目标航迹起始方法,通过引入一种泛化熵可调参数,解决了传统随机森林算法对差异数据集泛化能力和适应性差且难以确定全局最优解的难题。首先利用雷达多周期回波数据构建随机森林算法决策模型,其次基于设计的泛化熵处理规则对实测数据进行分类处理得到所需测试样本,最后将测试样本输入完成模型训练的随机森林进行迭代寻优和自主决策以获取雷达目标的航迹起始结果。雷达实测数据验证表明,文章所提出的方法相比于传统随机森林方法具有更为优越的目标航迹起始性能。     

The use of solutions to problems of spacecraft trajectory optimization in impulse formulation when solving the problems of optimal control of trajectories of a spacecraft with limited thrust engine: I

In this first part of our paper, it is suggested to use solutions to boundary value problems in the optimization problems (in impulse formulation) for spacecraft trajectories in order to obtain the initial approximation, when boundary value problems of the maximum principle are solved numerically by the shooting method. The technique suggested is applied to the problems of optimal control over motion of the center of mass of a spacecraft controlled by the thrust vector of jet engine with limited thrust in an arbitrary gravitational field in a vacuum. The method is based on a modified (in comparison to the classic scheme) shooting method computation together with the method of continuation along a parameter (maximum reactive acceleration, initial thrust-to-weight ratio, or any other parameter equivalent to them). This technique allows one to obtain the initial approximation with a high precision, and it is applicable to a wide range of optimal control problems solved using the maximum principle, if the impulse formulation makes sense for these problems.     

Computer search of variants for vehicle system composition synthesis

The design experience of vehicles systems composition synthesis is summerized. Some engineering techniques are suggested, which give an opportunity to solve the problems of a composition synthesis by means of a computer. These techniques are based on the following principles: construction of statistical models for the analysis of criteria behaviour; using of methods of the theory of extreme values statistics for the evaluation of the criteria limiting values; deriving of suboptimum variants groups; using of heuristic procedures for a cyclic correction of the search process in order to specify the range of boundary solutions for the problems with several criteria; using of a probabilistic appraisal for an estimation of search results; construction of educating algorithms using a fine for the variable values (composition variants); clearing up in the course of a dialogue process the interaction of an engineer with a computer.As in the design practice no requirements are extended to the algorithms of vehicle system characteristics design, that is why it is not possible at the stage of design solutions search to use the known recommendations worked out for the problems of linear, non-linear and dynamic programming.     

The Solution of Some Problems of Optimizing the Processes of a Flexible Correction of Motion of Satellite Systems: II.

In a previous paper [1], as a result of group-theoretic decomposition, one of the problems of the optimization of the flexible correction process was reduced to a series of subproblems. In this paper, the analytical solutions to these subproblems are obtained by the sweep method. A series of other similar problems with different variants of flexibility is also considered; the decomposition of these problems is performed, and their analytical solutions are presented. It is shown for some numerical examples that flexible corrections require essentially lower fuel consumption than the rigid correction, where all parameters characterizing the absolute motion of a satellite system are corrected to their nominal values.     

基于LMI的H-/H∞故障检测观测器设计

针对系统模型的不确定性、未知输入扰动,为提高对干扰鲁棒性、待检故障敏感性及降低设计的保守性,提出了一种H-/H∞故障检测观测器设计方法,通过无损procedure将非凸约束问题转化为线性矩阵不等式(LMI)约束,采用Schur补引理将H-/H∞故障检测观测器设计转化为凸优化问题的LMI表述;设计了自适应门限以减小故障检测过程中的误报、漏报率。故障检测观测器的设计是借助凸优化问题的数值计算,使设计过程便于实现。提出的算法使未知扰动的鲁棒性及待检故障的敏感性之间具有最优平衡。仿真验证表明,所提算法使残差具有快速的收敛性、较短的故障检测时间。