Chaotic patrol robot with frequency constraints

Abstract

Patrol or surveillance robots follow a trajectory that should be hard to predict. Dynamic nonlinear systems like chaotic systems are popularly used for generating pseudorandom trajectory for patrol robots. Due to apparently random nature of the trajectory, the presence of the robot at a certain region of the target area becomes infrequent. These point visit frequencies needs to be bounded or constrained so that a certain region is not left unmonitored for a long duration. In this paper we provide a proper survey of the issue. We also propose an approach that dynamically infuses separately generated chaotic coordinates into the original trajectory, such that the robot visits every significant part of the area at least once within a certain time. We also keep in check that the overall randomness of the system is not compromised.