Adaptive neuro-fuzzy expert systems for predicting specific energy consumption and energy stability margin in crab walking of six-legged robots

Abstract

In this paper, adaptive neuro-fuzzy expert systems have been designed to predict specific energy consumption and normalized energy stability margin for crab walking of a six-legged robot. The application of this technique for crab gait generation of the six-legged robot is new, to the best of the authors' knowledge. Three approaches based on adaptive neuro-fuzzy inference system have been developed and their performances are compared with each other. Genetic algorithm-tuned multiple adaptive neuro-fuzzy inference systems are found to perform better than other approaches. This could be due to a more exhaustive search carried out by the genetic algorithm compared to back-propagation algorithm and the use of two separate adaptive neuro-fuzzy inference systems for two different outputs.