A Study of Fuzzy-Neural Force Control for a Quadrupedal Walking Machine

Abstract

In this paper, the inverse kinematics and the foot force distributions of a planetary gear type quadrupedal walking machine are analyzed. This walking machine has four legs and each leg has three-degrees-of-freedom. The whole system has a total of seventeen links. The planetary gear leg is designed for a quadruped which can walk and trot under the following design criteria: high efficiency, compact size, and high payload/weight ratio. A neural network structure fuzzy logic controller is applied to control the foot force distribution of the quadruped walking machine. The performance of this fuzzy logic control algorithm is evaluated. Experimental results show that the fuzzy logic controller is effective in controlling this walking machine.