Title :
Towards distributed coverage of complex spatiotemporal profiles
Author :
Oyekan, John ; Hu, Huosheng
Author_Institution :
Sch. of Comput. Sci. & Electron. Eng., Univ. of Essex, Colchester, UK
Abstract :
Inspired by self-organization in natural organisms, an approach that would enable robotic agents form a visual representation of an invisible distributed hazardous substance is presented. Such a resource would enable humans observe and stay away from areas of high hazardous substance concentration. In this work, a proportional-integral control law and a machine learning scheme is used to obtain optimal parameter values that would enable optimal visual mapping whilst keeping computational resources low.
Keywords :
PI control; environmental factors; learning (artificial intelligence); robots; complex spatiotemporal profiles; distributed coverage; invisible distributed hazardous substance; machine learning scheme; natural organisms; optimal parameter; proportional-integral control law; robotic agents; Equations; Genetic algorithms; Machine learning; Mathematical model; Microorganisms; Pi control; Robots; Bio-inspired Algorithms; Environmental monitoring; Optimal Coverage; Self-Organization; Template learning;
Conference_Titel :
Robotics and Biomimetics (ROBIO), 2011 IEEE International Conference on
Conference_Location :
Karon Beach, Phuket
Print_ISBN :
978-1-4577-2136-6
DOI :
10.1109/ROBIO.2011.6181600
