Dikairono, Rudy (2021) Kecerdasan Buatan Untuk Persepsi, Formasi, Dan Perencanaan Gerak Robot Sepak Bola Beroda. Doctoral thesis, Institut Teknologi Sepuluh Nopember.
![[thumbnail of 07111660010007-Disertation.pdf]](http://repository.its.ac.id/style/images/fileicons/text.png) |
Text
07111660010007-Disertation.pdf - Accepted Version Restricted to Repository staff only until 1 October 2023. Download (3MB) | Request a copy |
|
Text
07111660010007-Disertation.pdf Restricted to Repository staff only Download (3MB) | Request a copy |
|
Text
07111660010007-Disertation.pdf Restricted to Repository staff only Download (3MB) | Request a copy |
|
Text
07111660010007-Disertation.pdf Restricted to Repository staff only Download (3MB) | Request a copy |
|
Text
07111660010007-Dissertation.pdf Restricted to Repository staff only Download (3MB) | Request a copy |
Abstract
Perkembangan robot sepak bola menjadi fokus banyak peneliti yang tergabung dalam konsorsium RoboCup. Salah satu bidang penelitian yang banyak diminati adalah robot sepak bola beroda. Pada disertasi ini dikembangkan robot sepak bola beroda dengan tiga tahapan penelitian. Penelitian pertama adalah pembuatan platform robot sepak bola beroda berupa desain mekanik dan elektronik yang mengacu pada aturan nasional dan internasional robot sepak bola beroda. Robot yang dibuat adalah robot Institut Teknologi Sepuluh Nopember Robot with Intelligent System (IRIS). Penelitian kedua adalah pembuatan sistem persepsi robot menggunakan Convolutional Neural Network (CNN), yaitu penambahan gambar omni-vision lapangan dengan pemodelan visual data kompas-gyro. Penelitian ketiga adalah pembuatan strategi baru untuk permainan robot sepak bola beroda. Strategi tersebut digunakan dalam permainan robot sepak bola beroda untuk penyerangan dan pertahanan. Strategi baru juga dikombinasikan dengan penetapan peran robot yang bisa berubah secara dinamis sehingga robot dapat berubah status dari robot penyerang menjadi robot bertahan dan sebaliknya. Hasil penelitian pertama adalah sebuah robot yang mampu bergerak mengikuti lintasan target yang direncanakan dengan RMSE sebesar 6,52 cm untuk lintasan lurus, RMSE 10.59 cm untuk lintasan melingkar, dan RMSE 18,42 cm untuk jalur lurus dengan hambatan. Hasil penelitian kedua adalah sebuah sistem persepsi berbasis CNN yang dapat meningkatkan akurasi lokalisasi diri robot di lapangan dari 32.8% menjadi 68.8%. Hasil penelitian ketiga adalah sebuah strategi baru untuk penyerangan dan pertahanan permainan robot sepak bola beroda yang telah diuji coba dengan hasil menang 80%, imbang 6,7%, dan memperoleh selisih gol tertinggi sebanyak 85 gol. Strategi penyerangan dan pertahanan baru ini juga menggunakan kontrol Fuzzy Inference System (FIS) dalam pengejaran dan penangkapan bola. Sistem ini mampu mengejar dan menangkap bola dengan sangat baik. Penelitian disertasi ini telah diimplementasikan dalam pertandingan Kontes Robot Sepak Bola Indonesia Beroda (KRSBIB) tahun 2019. Dalam pertandingan tersebut tim IRIS mendapatkan gelar Juara 1, mendapatkan gelar Desain Terbaik, dan mendapatkan gelar Strategi Terbaik.
=================================================================================================
The development of soccer robots has become the focus of many researchers who are members of the RoboCup consortium. One area of research that has been widely studied is the wheeled soccer robot. In this dissertation, a wheeled soccer robot was developed with three stages of research. The first research is the design and implementation of a wheeled soccer robot platform in mechanical and electronic that refer to national and international regulations. The robot is named Institut Teknologi Sepuluh Nopember Robot with Intelligent System (IRIS). The second research is the design and implementation of a robot perception system based on CNN. This perception system uses a Visual Modelling of a Gyrocompass Line Mark and Omni-vision Image for self-localization. The third research is the design and implementation of a new strategy for the wheeled soccer robot game. This strategy is used in the wheeled soccer robot game in attack and defense strategy. The new strategy is also combined with the dynamic role assignment that the robot can change the role from an attacking robot to a defensive robot and vice versa. The result of the first research was a robot that was able to move along the planned target trajectory with an RMSE of 6.52 cm for a straight line, RMSE of 10.59 cm for a circular trajectory, and RMSE of 18.42 cm for a straight line with obstacles. The second research result is a CNN-based perception system that can increase the robot's self-localization accuracy in the soccer field from 32.8% to 68.8%. The result of the third research is a new strategy for attack and defense of the wheeled soccer robot game that has been tested with 80% winning results, 6.7% draw, and the highest goal difference of 85 goals. This new attack and defense strategy also uses FIS control in chasing and catching the ball. This system shows excellent performance for chase and catches the ball. This dissertation research has been implemented in the 2019 Indonesian Wheeled Soccer Robot Contest (KRSBIB) competition. In this competition, the IRIS team won 1st place, got the Best Design title, and got the Best Strategy title.
| Item Type: | Thesis (Doctoral) |
|---|---|
| Uncontrolled Keywords: | convolutional neural network, formasi robot sepak bola , game play simulator, kerja sama robot, pemodelan visual, persepsi visual, robot sepak bola beroda, strategi menyerang, strategi bertahan, attacking strategy, convolutional neural network, defending strategy, gameplay simulator, robotic cooperation, robot soccer formation, visual modeling, visual perception, wheeled soccer robot. |
| Subjects: | T Technology > TJ Mechanical engineering and machinery > TJ211 Robotics. T Technology > TJ Mechanical engineering and machinery > TJ211.4 Robot motion T Technology > TJ Mechanical engineering and machinery > TJ211.415 Mobile robots T Technology > TK Electrical engineering. Electronics Nuclear engineering |
| Divisions: | Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Electrical Engineering > 20001-(S3) PhD Thesis |
| Depositing User: | Rudy Dikairono |
| Date Deposited: | 15 Aug 2021 04:20 |
| Last Modified: | 15 Aug 2021 04:20 |
| URI: | http://repository.its.ac.id/id/eprint/86744 |
Actions (login required)
 |
View Item |