【作者】 罗亚中；

【导师】 唐国金；

【作者基本信息】 国防科学技术大学 ， 航空宇航科学与技术， 2007， 博士

【摘要】 交会对接是进行空间组装、空间救援、深空探测等的必要技术,是我国载人航天第二步战略的关键技术。交会路径规划是交会对接技术中关键问题之一。论文以解决工程实际中的关键问题为目的,同时立足于拓展新的理论和方法,系统地研究了摄动和路径约束最优交会路径规划问题、多目标最优交会路径规划问题以及考虑偏差的鲁棒最优交会路径规划和实时交会路径规划问题。全文主要研究成果如下:发展了摄动和路径约束的最优交会路径规划策略。1)结合遗传算法的全局搜索能力和序列二次规划算法的快速收敛能力以及简化解析轨道计算和高精度轨道数值计算的优点,提出了一类求解考虑摄动和路径约束最优交会问题的两步串行混合优化策略;2)提出了结合主矢量理论、Lambert交会算法和进化算法用于确定最优脉冲数目的交互式规划策略;3)建立了调相交会特殊点变轨的整数和连续变量混合变量优化模型,提出了结合整数编码遗传算法和牛顿迭代法的混合求解策略,给出了不可行解剔除、记忆数据库以及迭代初始点更新三种用于提高算法鲁棒性和效率的措施;4)给出了一个新的多圈Lambert交会迭代算法,建立了基于多圈Lambert交会算法的调相交会综合变轨规划模型,提出了基于并行模拟退火单纯形算法的全局优化策略。基于解析问题和实际应用问题对所提出规划策略的全局收敛能力、效率和鲁棒性进行了充分的测试和验证。研究提出了多目标最优交会路径规划的理论、模型和算法。1)定义了交会轨道被动安全的一个评价指标,并给出了一类半解析分析算法;2)分别基于线性C-W交会方程和非线性二体交会方程,建立了综合考虑燃料消耗、交会时间和交会轨道被动安全指标等三个指标的多目标最优交会路径规划模型,采用一类多目标遗传算法NSGA-Ⅱ获得Pareto最优解集,优化结果可以快捷有效地揭示交会轨道整体性能指标之间的关系;3)提出了基于物理规划算法、Lambert交会算法和模拟退火算法的非线性摄动多目标最优交会问题的求解策略,获得了能应用于实际飞行任务并满足设计者偏好指标的满意解。研究提出了考虑偏差鲁棒最优交会路径规划的理论、模型和算法。1)定义了偏差最优交会的鲁棒评价指标,提出了一类用于非线性交会偏差分析的协方差分析方法;2)分别基于线性C-W交会方程和非线性二体交会方程,建立了综合考虑燃料消耗和鲁棒指标的鲁棒最优交会两目标规划模型,采用NSGA-Ⅱ获得了Pareto解集,通过与燃料最优解和鲁棒指标最优解的对比验证了解的Pareto最优性;3)将鲁棒指标引入到多目标交会路径规划中,优化结果可以快速地获得考虑偏差交会轨道整体性能指标之间的关系。研究给出了非线性交会实时路径规划算法。1)给出了基于Lambert算法的考虑轨道摄动的自主交会轨道制导算法;2)发展了调相交会的轨道机动实时规划算法,包括特殊点变轨实时规划算法和基于近圆偏差方程的综合变轨实时规划算法;3)研制了集成论文研究成果并面向工程应用的交会路径规划软件。论文研究拓展了现有最优交会路径规划的研究范畴,发展了多目标最优交会路径规划和考虑偏差的鲁棒最优交会路径规划的理论、模型和算法,具有一定的理论意义,同时为交会轨道的综合性能指标设计提供了新的手段;论文所提出的复杂交会路径规划方法、实时路径规划算法以及所研制的交会路径规划软件解决了实际工程应用的难点问题,在我国载人航天二期工程和其他空间应用工程中具有较高的实践价值。更多还原

【Abstract】 Rendezvous and docking technology is essential in space-assembly, space-succor, outer space-exploration etc., and is a key technology for the second step of China manned spaceflight project. Rendezvous trajectory planning is a critical problem in rendezvous and docking technology. For the purpose of solving the critical engineering application problems and further developing the new theory and methods, this dissertation studies the optimal rendezvous trajectory with perturbations and path constraints planning problem, the multiobjective rendezvous trajectory planning problem, the robust optimal rendezvous trajectory planning problem and the real-time trajectory planning problem with considering uncertainties. The main results achieved in this dissertation are summarized as follows.The optimal planning approaches for rendezvous trajectory with trajectory perturbations and path constraints are developed. 1) A hybrid approach combining the global search ability of genetic algorithm and the high convergence rate of SQP, together with the simplified analytical propagator and the high-fidelity mathematical propagator is proposed to solve the perturbed and path-constrained multi-impulse fuel-optimal rendezvous problem. 2) An interactive optimization approach using the evolutionary algorithm, the Lambert rendezvous algorithm and the Primer vector theory is proposed to optimize a rendezvous trajectory with the optimal number of impulses. 3) The design of a rendezvous phasing strategy with special-point maneuvers is formulated as a mixed integer and continuous nonlinear programming problem. A hybrid approach combining an integer-coded genetic algorithm with Newton’s method is proposed for solving this problem. Three improvements including the infeasible solution return, the memory database and renewing the initial guess are imposed on the hybrid approach to make it more efficient and robust. 4) A new multiple-revolution Lambert algorithm is proposed, and an optimization model is formulated for the design of a rendezvous phasing strategy with combined maneuvers, and a global optimization approach based on a parallel simulated annealing using simplex method is proposed. The global convergence ability, high efficience and robustness of these four proposed optimization approaches are testified by solving the analytical problems and the practical problems.The optimal multiple-objective rendezous trajectory planning theory, methods and approaches are studied and proposed. 1) A performance index related to rendezvous trajectory passive safety is defined and a semi-theoretical model for calculating this safety performance index is provided. 2) Based on the Clohessy-Wiltshire linearized rendezouvs equations and two-body Lambert rendezvous equations, the three-objective optimization models including the minimum characteristic velocity, the minimum time of flight and the maximum safety performance index are respectively formualted, the nondominated sorting genetic algorithm(NSGA-Ⅱ) is employed to obtain the Pareto solution set. The optimizaiton results can quickly demonstrate the tradeoffs between the whole performace indexes of a rendezvous trajectory. 3) The optimization approach using the physical programming, the Lambert rendezvous algorithm and the simulated annealing for the multi-objective optimization design of nonlinear perturbed rendezvous is proposed, the designer preferred solution which can be directly applied to mission plan is obtained.The robust optimal rendezous trajectory with uncertainties planning theory, methods and approaches are studied and proposed. 1) One robustness performance index of a rendezvous trajectory with uncertainty is defined, and one covariance analysis method for nonlinear rendezvous uncertainty analysis is proposed. 2) Based on the Clohessy-Wiltshire linearized rendezouvs equations and the two-body rendezvous equations, a two-objective optimization model including the minimum characteristic velocity and the minimum robustness performance index are formuated, the NSGA-Ⅱis employed to obtain the Pareto solution set. The Pareto optimality of the obtained solutions is testified by comparing with the fuel-optimal and robustness-optimal solutions. 3) The robustness performance index is introduced into the multi-objecitve rendezvous trajectory planning. The results can quickly demonstrate the tradeoffs between the whole performance indexes of a rendezvous trajectory with uncertainty.The realtime rendezous trajectory planning approaches are studied and provided. 1) The trajectory guidance algorithm which considers trajectory perturbations for autonomous rendezvous is proposed based on the Lambert algorithm. 2) The real-time planning approach for phasing rendezvous trajectory maneuvers, including the special-point maneuvers real-time planning approach and the combined maneuvers real-time planning approach using the near circular deviation equations, are developed. 3) A rendezvous trajectory planning software which integrates the theoretical achievements of this dissertation is developed for engineering application.This dissertation expands the research domain of the current rendezvous trajectory planning by developing the optimal multiple-objective and robust optimal rendezous trajectory planning theory, methods and approaches, which have some theoretical significance and also provide new design tools for comprehensive performace index optimization design of a rendezvous trajectory. The proposed complex rendezvous trajectory planning approach, the real-time rendezvous planning approach and the developed rendezvous trajectory planning software solve the ctritical engineering problems, which have high application value in China manned spaceflight project and other space application projects.更多还原