作者: K. Cerbioni , C. Colosimo , A. Starita , P. Dario
DOI: 10.1007/978-3-7908-1859-8_28
关键词: Fuzzy logic 、 Six degrees of freedom 、 GRASP 、 Fuzzy control system 、 Artificial intelligence 、 Representation (mathematics) 、 Inference engine 、 Engineering 、 Neuro-fuzzy 、 Emulation
摘要: In this paper a neuro-fuzzy system, which realises high, level control of the pre-grasping movements, in an advanced sensorised robotic manipulator, is presented. The proposed system based on modelling human grasp order to transfer capabilities new systems. made two subsystems: neural module and fuzzy one, fully integrated, assuming information about geometric features describing objects are given motor by artificial vision module. tool builds knowledge base their representation, producing rules that are, successively, passed inference engine gives, as output, right configuration hand chosen object. modules can change increase with experience, incrementally. has been implemented equipped six degrees freedom (d.o.f.) arm artificial, d.o.f., hand. Different objects, varied spatial orientations, have used develop different tasks.