การตัดสินใจแบบฟัซซีสำหรับหุ่นยนต์ผู้รักษาประตู

Control system and decision making in soccer robot are complex. This leads to an uncertain condition definition and incorrect robot command selection. In this work, we develop a system that controls goal keeper robot using Mamdani fuzzy inference system. In particular, the system will make a decisio...

Full description

Saved in:
Bibliographic Details
Main Author: ประเสริฐ ฉินสวัสดิ์พันธุ์
Other Authors: ผู้ช่วยศาสตราจารย์ ดร.ศันสนีย์ เอื้อพันธ์วิริยะกุล
Format: Independent Study
Language:Thai
Published: เชียงใหม่ : บัณฑิตวิทยาลัย มหาวิทยาลัยเชียงใหม่ 2017
Subjects:
Online Access:http://cmuir.cmu.ac.th/jspui/handle/6653943832/39963
Tags: Add Tag
No Tags, Be the first to tag this record!
Institution: Chiang Mai University
Language: Thai
Description
Summary:Control system and decision making in soccer robot are complex. This leads to an uncertain condition definition and incorrect robot command selection. In this work, we develop a system that controls goal keeper robot using Mamdani fuzzy inference system. In particular, the system will make a decision on robot movement to protect a goal from assessing football position in the field. The expert’s opinion is utilized to help in defining a suitable system structure, number of linguistic variables, membership functions, and rules. There are 3 experiments to show the efficiency of Fuzzy decision for goal keeper robot. The first experiment is to measure the accuracy of position assessment of football position in the field. The accuracy of identifying a correct area in the field is 95.58% with only 4.42 % incorrect identification. The system is able to assess the football position correctly and that leads to a suitable controlled robot movement techniques, i.e., Followball or Intercepball. The fuzzy decision making system for goalkeeper robot efficiency is shown in the second experiment. We implement a goal protecting system with penalty shooting in 40 different positions with 2000 shooting at each position. The result shows that the system can protect a goal 99.9%. We implement the fuzzy decision making system for goalkeeper robot in a simulated game situation in the third experiment. There are 50 simulated game situations with 3 minutes in each game. The system is able to protect the goal 98.97%. From the study, our system is able to make a good decision and able to assess varied situations in the field. The goalkeeper robot is able to protect the goal from opponent. However, to help improve the system, we need to select proper membership functions.