Abstract
Complex behaviour generation for artificial systems requires huge implementation efforts that can be partially overcome by decomposing global tasks in simpler well specified behaviours easier to design and to tune independent of each other. Robots behaviour can be implemented as a set fuzzy rules that mimic expert knowledge in specific tasks as a way to model expert knowledge. Present work shows a complex navigational behaviour as it is Safe follow wall that can be decomposed in three basic behaviours. Two combination lnethods are proposed and compared; they are controlled by fuzzy rnetarules to fuse the actions derived from the predefined elementary behaviours. These actions combination/arbitration algorithms have been demonstrated in both simulated and real worlds. Some results are presented to display the major issues concerned to each of the proposed algorithms.
Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
