基于SQP方法的常推力月球软着陆轨道优化方法

月球软着陆是未来月球探测中的一项关键技术。针对这项技术，本文给出了一种基于SQP方法的常推力月球软着陆轨道优化方法。该方法通过将常推力月球软着陆轨道离散化，利用离散点处状态连续作为约束条件，把常推力月球软着陆轨道优化问题归结为一个非线性规划问题，对于此问题的求解，其初值均为有物理意义的状态和控制量，从而避免了采用传统优化方法在解决此优化问题时对没有物理意义变量初值的猜测。最后，利用SQP方法求解了此轨道优化问题。仿真计算结果表明这种离散化的方法应用于此轨道优化问题可以避免传统轨道优化方法对初值敏感的问题。

Study on the Optimal Trajectories of Lunar Soft-Landing with Fixed-Thrust Using SQP Method

The Lunar Soft-landing will be the significant technique of lunar exploration mission in the future. To this technique, an optimization approach of lunar soft-landing with fixed-thrust using SQP method is presented. It made the lunar soft-landing trajectories with fixed-thrust to be discrete. At the discrete point, the continuity of motion states was regarded as the constraint conditions, then the problem of the lunar soft-landing trajectories with fixed-thrust could be translated into a nonlinear program problem. During solving this problem, all of the initial values, including the motion states and control value, had the physical meanings. So the guessing the initial values, which had no physical meanings, was avoided. Finally, it solved this problem using Sequential Quadratic Programming (SQP) method. The results of calculation and simulation indicate that application of the discrete method for problem of trajectory optimization can avoid the problem of initial value sensitivity.