Implementasi Kontrol Formasi Kooperatif Leader–Follower pada Tiga Yahboom RDK X5 Menggunakan Algoritma DWA untuk Pelacakan Jalur Menuju Target

Ramadhinta, Nikolaus Rangga (2026) Implementasi Kontrol Formasi Kooperatif Leader–Follower pada Tiga Yahboom RDK X5 Menggunakan Algoritma DWA untuk Pelacakan Jalur Menuju Target. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5022221164-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (10MB)

Abstract

Penelitian ini mengimplementasikan kontrol formasi kooperatif pada sistem multirobot leader-follower menggunakan algoritma Dynamic Window Approach (DWA). Platform yang digunakan adalah tiga robot Yahboom RDK X5 berpenggerak roda mecanum dengan sensor LiDAR, IMU, dan encoder roda, serta lokalisasi AMCL pada peta lingkungan yang telah dibangun sebelumnya. DWA berperan sebagai perencana lokal yang mengoptimalkan pemilihan kecepatan linear dan sudut dengan mempertimbangkan batasan kinematis dan jarak aman terhadap rintangan. Robot leader menavigasi menuju target, sedangkan kedua robot follower mempertahankan jarak dan sudut relatif terhadap leader menggunakan kontrol PID dengan feedforward dan pertukaran informasi antar robot melalui komunikasi UDP. Sistem juga mendukung formasi dinamis berupa perubahan bentuk formasi secara otomatis saat berjalan, serta modul Fault Generator yang menyuntikkan gangguan acak pada follower untuk menguji ketahanan, dengan pemulihan formasi secara kooperatif melalui pengatur kecepatan pada leader. Seluruh implementasi dan pengujian dilakukan langsung pada tiga robot fisik menggunakan Robot Operating System 2 (ROS 2). Evaluasi kinerja mencakup error formasi jarak dan sudut, kecepatan pembentukan dan perpindahan formasi, ketelitian lintasan leader terhadap jalur rencana, serta waktu pemulihan saat terjadi gangguan. Hasil pengujian menunjukkan sistem mampu bergerak secara kooperatif, mempertahankan formasi dengan error kecil, berpindah bentuk dengan respons cepat tanpa tabrakan antar robot, menjejak target secara efektif, dan memulihkan formasi setelah gangguan, dengan ketelitian lokalisasi sebagai faktor pembatas utama kinerja.
======================================================================================================================================
This study implements cooperative formation control in a leader-follower multirobot system using the Dynamic Window Approach (DWA) algorithm. The platform used is three Yahboom RDK X5 robots with mecanum wheels with LiDAR sensors, IMUs, and wheel encoders, as well as AMCL localization on a pre-built environmental map. DWA acts as a local planner that optimizes the selection of linear and angular velocities by considering kinematic constraints and safe distances to obstacles. The leader robot navigates towards the target, while the two follower robots maintain their distance and angle relative to the leader using PID control with feedforward and information exchange between robots via UDP communication. The system also supports dynamic formations in the form of automatic formation changes while walking, as well as a Fault Generator module that injects random disturbances into the followers to test resilience, with cooperative formation recovery through a speed controller on the leader. All implementations and tests were carried out directly on three physical robots using Robot Operating System 2 (ROS 2). Performance evaluations included distance and angle formation errors, formation formation and displacement speeds, leader trajectory accuracy relative to the planned path, and recovery time following disturbances. Test results demonstrated the system's ability to move cooperatively, maintain formation with minimal error, rapidly shift shape without collisions, effectively track targets, and recover formation after disturbances, with localization accuracy being the primary performance limiting factor.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Leader-Follower, Dynamic Window Approach (DWA), Yahboom RDK X5,Formation Control, Path Tracking, Fault Generator
Subjects:	T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK3070 Automatic control T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK6592.A9 Automatic tracking.
Divisions:	Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Electrical Engineering > 20201-(S1) Undergraduate Thesis
Depositing User:	Nikolaus Rangga Ramadhinta
Date Deposited:	24 Jul 2026 08:03
Last Modified:	24 Jul 2026 08:03
URI:	http://repository.its.ac.id/id/eprint/136966

Actions (login required)

View Item