Fuzzy Control of Underwater Walking Robot During Obstacle Collision Without Pre-defined Parameters

摘要

Industrial exploration of sea bottom requires special underwater technical devices, which moved at sea bottom. Walking machines in seabed conditions, which characterized by sophisticated shape and low bearing ability of the ground, have higher traction properties and possibility, compared with traditional vehicles. In the paper discussed some results of investigation of controlled movement dynamics of walking robot with cyclic mover during obstacle collision without pre-defined parameters. At the base of integration of information from video sensors, was offered situational method to organization behavior of the robot in conditions of incomplete and ambiguous understanding of the current situation and workspace. At the base of fuzzy movement control algorithms were developed typical leg motions, autonomously performed by the robot with the purpose of exclusion of emergency situations. Interactive analysis of the device control system as software model of control object showed, that majority of typical situations on the control can be solved without operator intervention. Experimental verification of gained results of mathematical modeling at the base of underwater walking robot MAK-1 was held. During underwater experiments dynamics of walking device, traction properties and possibility were investigated. In research of possibility were passed local obstacles both straight movement mode and in special maneuvering mode. Some attention during the tests was given to tuning of the robot fuzzy control system in training mode. Results of the work can be demand in underwater walking robotic systems development for underwater technical works and for new industrial technologies of seabed resources reclamation.