Adaptive navigation control for a quadruped robot based on ART and fuzzy logic

An adaptive navigation control problem is presented for a quadruped robot in cluttered environments. by incorporating the capability of adaptive resonance theory (ART) in stable category recognition into fuzzy logic. control. An ART-based neural network, is introduced as an environment identifier for the purpose of adaptive selection of the adequate rule base for a fuzzy controller. Therefore. the proposed adaptive control scheme for the navigation of the robot is implemented by the adaptive fuzzy rule base in response to changes of the robot's environment. which can be fine observed by the proposed environment identifier. Some simulation, results are presented to illustrate the proposed algorithm.