基于解析梯度的近地小天体探测机会搜索问题研究

Search of Exploration Opportunity for Near Earth Objects Based on Analytical Gradients

研究近地小天体的探测机会搜索问题.针对交会型探测任务,通过结合变分理论和状态转移矩阵推导了性能指标关于可调参数的解析偏导数,然后在搜索空间中随机生成初始点,并从这些初始点出发利用解析偏导数寻优,从而得到搜索空间内对应潜在发射机会的全部局部极小值点.此方法既在一定程度上保持了传统搜索方法全局搜索的特点,又克服了传统搜索方法的盲目性,因此计算速度获得很大提高.此外该方法可以对探测机会的搜索精度进行有效地控制.

The problem of searching for exploration opportunity of near Earth objects is investigated. For rendezvous missions, the analytical gradients of performance index with respect to free parameters are derived by combining the calculus of variation with the theory of state-transition matrix. Then, some initial guesses are generated random in the search space, and the performance index is optimized with the guidance of analytical gradients from these initial guesses. This method not only keeps the property of global search in traditional method, but also avoids the blindness in the traditional exploration opportunity search; hence, the computing speed could be increased greatly. Furthermore, by using this method, the search precision could be controlled effectively.