Fuzzy logic controlled autonomous quadruped robot

Abstract

Researchers have taken several steps to build an appropriate 4-point based quadruped robot controller. These efforts can be categorized into analytical and intelligent methods into two main fields. The study of an analytical method is based on the robot’s primarily complex movement equations. On the other hand, an intelligent method requires less knowledge of kinematic and dynamic equations to control the motion and orientation of the quadruped robot autonomously in any environment. In this study, we have chosen the Fuzzy Controller (FC) as an intelligent sensor-actuator control system for the quadruped robot. This robot includes sensors to sense the physical quantities of the environment. And the outputs of the sensors are used as inputs to the FC. According to the sensor inputs and defined rule-based system, the controller gives the actuator control commands to the robot's leg. Further on, the fuzzification and defuzzification modules make decisions that control the robot's motion and allow the robot to navigate automatically. The two-dimensional and three-dimensional models of the quadruped robot have been designed through the SolidWorks software, and the fuzzy controlled simulations of the quadruped robot are presented in the Virtual Robot Experimentation Platform (V-REP) software.