Abstract A non-holonomic constant-speed robot travels in an unknown maze-like environment cluttered with complex obstacles. Through the obstacle-free part of the plane, the robot should autonomously arrive at the isoline where an unknown scalar field assumes a given value. Afterwards, it should track the obstacle-free part of the isoline. The robot has access only to the field value at the current location and the distance from this location to the obstacles. We present a hybrid nonlinear navigation law that solves this mission. The law does not use estimation of the field gradient and is non-demanding with respect to both computation and motion. The non-local convergence of the proposed algorithm is rigorously justified and confirmed by computer simulation tests.

中文翻译：

---

非完整机器人在类似迷宫的环境中跟踪环境水平集的混合控制

摘要 一个非完整恒速机器人在一个未知的迷宫般的环境中行走，环境复杂，障碍物杂乱无章。通过飞机的无障碍部分，机器人应该自主地到达等值线，其中未知标量场具有给定值。之后，它应该跟踪等值线的无障碍部分。机器人只能访问当前位置的字段值以及从该位置到障碍物的距离。我们提出了一种混合非线性导航法则来解决这个任务。该定律不使用场梯度的估计，并且对计算和运动都没有要求。所提出算法的非局部收敛性得到了计算机模拟测试的严格证明和证实。