A Multi-Objective Exploration Strategy for Mobile Robots

Exploration strategies are used to guide mobile robots in building maps of environments. Usually, exploration strategies work greedily by evaluating a number of candidate observation positions on the basis of a utility function and selecting the best one. The utility functions are defined in an ad hoc manner as the compositionof values measuring different features of a candidate observation position, such as the travelling cost and the estimated information gain. In this paper, we propose a more general way to define an exploration strategy through multi-objective optimization. In our approach, the values of the features are kept separated without combining them in a particular utility function. Experimental results demonstrate the effectiveness of our method.