Multiobjective aerial surveillance over disjoint rectangles
Multiobjective aerial surveillance over disjoint rectangles
复制标题
DOI：
10.1016/j.cie.2020.106732
复制
发表时间：
2020-10
期刊：
影响因子：
--
通讯作者：
Orhan Karasakal;E. Karasakal;Güliz Maraş
中科院分区：
文献类型：
--
作者：
Orhan Karasakal;E. Karasakal;Güliz Maraş
在空中监视问题（ASP）中，一个装有监视传感器的空中平台通过覆盖多个矩形区域，然后返回到它开始的地方。在本文中，我们提出了一个多目标扩展ASP，其目的是帮助航空使命规划者达到他/她的最佳解决方案中的一组有效的替代品。我们考虑两个相互冲突的目标，最大限度地减少行驶距离和最大限度地提高目标检测的最小概率。每个目标都可以用来解决单目标ASP。然而，从使命规划者的角度来看，需要同时优化这两个目标。为了使使命规划者在冲突目标下能得到最优解，提出了精确和启发式的多目标ASP（MASP）方法。我们还开发了一个交互式的过程，以帮助使命规划者选择最满意的解决方案中的所有帕累托最优解。计算结果表明，所提出的方法使使命规划者能够在大量备选方案中捕获冲突目标之间的权衡，并在少量迭代中消除不需要的解。
In Aerial Surveillance Problem (ASP), an air platform with surveillance sensors searches a number of rectangular areas by covering the rectangles in strips and turns back to base where it starts. In this paper, we present a multiobjective extension to ASP, for which the aim is to help aerial mission planner to reach his/her most preferred solution among the set of efficient alternatives. We consider two conflicting objectives that are minimizing distance travelled and maximizing minimum probability of target detection. Each objective can be used to solve single objective ASPs. However, from mission planner’s perspective, there is a need for simultaneously optimizing both objectives. To enable mission planner reaching his/her most desirable solution under conflicting objectives, we propose exact and heuristic methods for multiobjective ASP (MASP). We also develop an interactive procedure to help mission planner choose the most satisfying solution among all Pareto optimal solutions. Computational results show that the proposed methods enable mission planner to capture the trade-offs between the conflicting objectives for large number of alternative solutions and to eliminate the undesirable solutions in small number of iterations.