Estimation Method for Operational Environment Complexity by a Robotic Team.

The paper presents the estimation method for operational environment complexity by a robotic team. Its distinguishing feature is the way to organize the intragroup information exchange between the team members, identifying the robots that, due to some restrictions, are not able to adequately estimate the complexity of the surrounding area. Overcoming the restrictions implies assistance from group members, meanwhile, their reasons may be of different origin, for example, faulty individual sensors or sensors’ visibility range being overlapped with environmental elements. Upon the method application, an unambiguous (explicit) idea concerning the information sources and receivers and their necessary list, is formed, which makes the exchange of data more efficient and the estimation process more accurate. A special function set has been developed for environment complexity estimation; using this function set implies representing the surrounding area as a plane divided into sectors with their dimensions corresponding to the team robot size. The transition from three-dimensional to two-dimensional space is carried out by projecting onto a flat surface the coordinates of the environment elements. The presented function set is approximated with high precision by an artificial neural network that can be used for further evaluations.