Humanoid robot path planning with fuzzy decision-making processes and deep learning techniques

نوع: Type: thesis

مقطع: Segment: masters

عنوان: Title: Humanoid robot path planning with fuzzy decision-making processes and deep learning techniques

ارائه دهنده: Provider: Erfan Saghabashi

اساتید راهنما: Supervisors: Dr. Hassan Khotanlou

اساتید مشاور: Advisory Professors:

اساتید ممتحن یا داور: Examining professors or referees: Dr. Reza Mohammadi & Dr MirHossein Dezfoulian

زمان و تاریخ ارائه: Time and date of presentation: 021/11/20

مکان ارائه: Place of presentation: Online

چکیده: Abstract: Route planning and obstacle detection for robots is important in robotics, especially in anthropomorphic robots, and is subject to many errors. Choosing the right path by the robot as well as recognizing obstacles are the most important steps in robot route planning. Recently, the development of deep learning techniques, especially in the field of machine vision, in this field has become one of the most important research fields, especially in the field of robotics, and it is very important that automation of the routing process is possible. Many robot path programming algorithms usually follow a local map or use only sensor information that does not respond well to changes. Due to the need to implement on a humanoid robot, Nao robot has been used to analyze the method in this study, which due to the unavailability of the robot itself, a simulated version has been used. In this research, different methods of combining deep learning and fuzzy system were tested to obtain the best combination. Robot path planning in this research has been done by combining object recognition using deep learning techniques and fuzzy decision making system based on rules. Research Implementation of the method This research is based on the NEO robot in a simulation environment and the robot can move around the environment. The result of learning and testing the proposed method in this study, using quantitative and qualitative criteria in comparison with previous methods was found that the combination of this method with fuzzy decision making system, leaves acceptable results and also the performance of the proposed method results. The path planning has significantly improved the previous methods

فایل: ّFile: Download فایل