论文部分内容阅读
应用Monte-Carlo法和遗传算法的联合仿真求解Lambert转移中途修正的全局概率最优策略.首先推广限制性三体问题中求解周期性特解的微分修正算法构造出考虑J2项摄动下的Lambert转移轨道并以此作为参考轨迹,则中途修正策略仅需针对导航误差、初始偏差修正的控制偏差等进行补偿.应用微分修正算法导出的单值矩阵,设计出3类线性和非线性中途修正策略,以适应不同的精度需要.随后应用Monte-Carlo和遗传算法的联合仿真,可以得到实现代价函数(落点误差最小)在概率意义下的最优解.与直接利用优化算法寻优需要已知各种误差量不同,得到的最优修正策略更具有普适性.
The co-simulation of Monte-Carlo method and genetic algorithm is used to solve global probabilistic optimization strategy of Lambert shift mid-way modification.Firstly, a differential correction algorithm for solving periodic unique solutions of the restricted three-body problem is proposed to construct Lambert Transfer trajectory and use it as a reference trajectory, the midway correction strategy only needs to compensate for the navigation error, the control deviation of the initial deviation correction, etc. Using the single value matrix derived from the differential correction algorithm, three kinds of linear and nonlinear midway correction strategies , To meet different needs of precision.Secondly, with the co-simulation of Monte-Carlo and Genetic Algorithm, we can get the optimal solution of the cost function (with the least error of placement) in the sense of probability.With the direct optimization algorithm optimization need to know The amount of different errors, the optimal correction strategy obtained is more universal.