An Improved Fuzzy Brain Emotional Learning Model Network Controller for Humanoid Robots.

摘要: The brain emotional learning (BEL) system was inspired by the biological amygdala-orbitofrontal model to mimic high speed of mechanism in mammalian brain, which has been successfully applied many real-world applications. Despite its success, such often suffers from slow convergence for online humanoid robotic control. This paper presents an improved fuzzy BEL (iFBEL) neural network integrating a (FNN) conventional BEL, effort better support robots. In particular, inputs are passed into sensory and channels that jointly produce final outputs network. non-linear approximation ability iFBEL is achieved taking as channel. proposed works with robust controller generating hand gait motion robot. updating rules iFBEL-based composed two parts, including channel followed model, FNN derived "Lyapunov" function. experiments on three-joint robot manipulator six-joint biped demonstrated superiority reference proportional-integral-derivative cerebellar articulation controller, based more accurate faster control performance iFBEL.