Chaotic Agent Navigation: Achieving Uniform Exploration Through Area Segmentation

A chaotic navigation algorithm is designed for an autonomous agent, that has the objective of exploring a given area, whilst moving unpredictably. The algorithm uses the Renyi map as a randomness source to generate the orientation of the robot and the direction it can move. There are eight possible directions, but the motion is limited to only three directions each time, which are determined by its orientation, so as to make the motion smoother and applicable to a real robot. To improve coverage, the area is segmented into sixteen equal subareas and the agent distributes its motion equally in each one, moving sequentially in each subarea. Two approaches were considered for the threshold of moving from one subarea to another. In the first, a move to an adjacent subarea is performed based on the number of executed steps. In the second, the move was performed after a certain coverage percentage is achieved. The simulations for the segmented area showed a slight rise in the coverage percentage, and a more consistent coverage across the whole area, when compared with the unsegmented technique.