Probabilistic Communication Based Potential Force for Robot Formations: A Practical Approach

LinkMind: Link Optimization in Swarming Mobile Sensor Networks

“…The robot-centered potential force represents the local influence of potential force to other robots and the environment [ 3 , 6 , 10 , 11 ]. The forces are usually based on local perception and/or inter-communication in which the attractive force pulls the robots close together while the repulsive force pushes the robots away from obstacles or other robots.…”

“…The work in [ 15 ] describes an approach to using exponential function to model attractive and repulsive forces. Our previous work [ 11 ] is also an example of the use of exponential function as potential force. It demonstrated that the developed forces are successful in keeping the robots in formation.…”

A review of swarm robotics tasks

“…The authors demonstrated with computer simulations that starting from particles in random locations the system can converge to a state where particles form a hexagonal lattice. Mikkelsen et al [125] used a similar virtual forcebased approach to obtain hexagonal lattice formations; their model includes the effects of communication errors between neighboring robots which occur due to the limited range and non-uniform coverage of the infrared sensors. Another study utilizing virtual forces is the work by Mathews [118], where formation of triangular grids is obtained; this pattern is shown to optimize area coverage by robots with given sensing and communication ranges, provided that the communication radius is sufficiently large compared to the sensing radius.…”

“…Communication can take place using explicit messages [115,116,117,118,119], or implicitly, by sensing the nearby presence and relative position of other robots [106,115,120,121,122,123,124,125,126,127]. Information acquired from nearby robots can be used to implement simple mechanisms of robot avoidance [123], or, more often, to regulate the position and velocity of a robot according to a desired behavior.…”

“…To evaluate the quality of the equilibrium state, especially in pattern formation techniques, some useful measures are the distribution of inter-robot distance [125] and the mean square error between desired and actual distance [119].…”