In the last years, a growing number of challenging applications in navigation, logistics, and tourism were modeled as orienteering problems. This problem has been proposed in relation to a sport race where certain control points must be visited in a minimal time. In a certain kind of these competitions, the choice of the number and the order for the control points are left to the competitor. We propose an original approach to solve the orienteering problem based on hybrid control. The continuous state of the system contains information about the navigation toward the next control point. In contrast, the discrete state keeps track of the already visited ones. The control problem is solved using non-standard dynamical programming techniques.
翻译:在过去的几年里,在航行、物流和旅游业中,越来越多的具有挑战性的应用被模拟为定向问题。 这个问题是在一场运动比赛中提出来的,在比赛中,必须在最短的时间内访问某些控制点。 在某种竞赛中,控制点的数量和顺序的选择由竞争者决定。 我们提出了一个基于混合控制的解决定向问题的原始方法。 系统的连续状态包含向下一个控制点的导航信息。 相反,离散状态会跟踪已经访问过的控制点。 控制问题通过非标准动态程序技术解决。