Loading...

Control of Quadruped Robot Using Fuzzy Finite State Machine

Jabbarifar, Mohammad | 2012

432 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 43595 (08)
  4. University: Sharif University of Technology
  5. Department: Mechanical Engineering
  6. Advisor(s): Bagheri Shouraki, Saeed; Meghdari, Ali
  7. Abstract:
  8. Nowadays, by increasing the applications of robots in military, industry and also home usages, researchers are looking for the most efficient algorithms to control the robots. As Prof. Zadeh presented the fuzzy systems in 1965 as a proper probabilistic logic, fuzzy control has been developed and now is one of the most applicable procedures in new aspects of controlling because of its less required computational capacity and more proper stability. Although the human like and quadruped robots are more applicable because of their coincidence with more difficult conditions compared to the wheeled robots, more complicated methods should be implemented to control them.The main subject of this study is decreasing the costs and complicacy of the robot controlling in different conditions by implementing the fuzzy methods to control the simulated quadruped robot.Finite States Machine (FSM) is a computational model of the machine with an initial internal memory which defines some finite states for the machine. Firstly, the initial state would be defined for the system and by changing inputs the state of the system will be changed based on the model. In Fuzzy Finite States Machine (FFSM), this procedure is more complicated. In fuzzy concepts, the system does not stand thoroughly in a defined state rather has the membership levels versus different states and based on these membership levels, the next state of the system should be defined.In this study, the basic concepts of quadruped robots and fuzzy systems are investigated afterwards the robot is simulated in WEBOTS. After selecting an appropriate gate for moving, the proper parameters for movement of the robot on different surfaces are extracted. Furthermore membership functions and fuzzy logics are resulted from different experiments. By substituting these properties in described controlling model, the robot can walk on different surfaces with different slope and friction coefficient.
  9. Keywords:
  10. Finite State Machine (FSM) ; Quadruped Robot ; Fuzzy Finite State Machine (FFSM) ; Movement Gate

 Digital Object List

 Bookmark

No TOC