Perencanaan Jalur Dinamis Untuk Navigasi Nonholonomic Mobile robot Menggunakan A* Modifikasi Dan Fuzzy Logic

Ikhsan, Muhammad (2025) Perencanaan Jalur Dinamis Untuk Navigasi Nonholonomic Mobile robot Menggunakan A* Modifikasi Dan Fuzzy Logic. Other thesis, Institu Teknologi Sepuluh Nopember.

Text 5022211075-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (3MB) | Request a copy

Abstract

Penggunaan mobile robot telah berkembang pesat dalam berbagai bidang, namun masih menghadapi tantangan signifikan dalam perencanaan jalur dan penghindaran rintangan di lingkungan dinamis karena algoritma A* tradisional memiliki kelemahan utama berupa waktu komputasi yang lama, menghasilkan jalur dengan banyak belokan tajam yang sulit diikuti mobile robot, dan ketidakmampuan menangani rintangan dinamis secara real time. Penelitian ini mengusulkan solusi komprehensif berupa algoritma A* modifikasi yang dikombinasikan dengan pendekatan fuzzy logic, meliputi pembatasan ekspansi neighborhood node hanya pada arah relevan, penambahan fungsi turn penalties untuk meminimalkan belokan tajam, path smoothing menggunakan G² Continuous Bezier Spiral (G2CBS), dan implementasi fuzzy logic kontroller untuk penghindaran rintangan dinamis. Hasil pengujian menunjukkan peningkatan performa signifikan dengan A* modifikasi berhasil mengurangi waktu komputasi hingga 50% (dari 13.7065s menjadi 6.8971s), mengurangi node ekspansi hingga 21% (dari 70.255 menjadi 55.443 node), mengurangi jumlah belokan hingga 86.30% (dari 73 menjadi 10 belokan), dan meningkatkan akurasi path tracking dengan RMSE turun hingga 82.63% (dari 0.0182 menjadi 0.0033) ketika menggunakan pure pursuit kontroller. Sementara untuk sistem penghindaran rintangan dinamis, mobile robot berhasil menghindari tabrakan pada jarak aman 8.83cm (obstacle 1) dan 7.33cm (obstacle 2) serta menghasilkan RMSE sebesar 0.929 yang menunjukkan keseimbangan optimal antara keselamatan dan akurasi pelacakan jalur.
===================================================================================================================================
The use of mobile robots has rapidly developed across various fields, yet still faces significant challenges in path planning and obstacle avoidance in dynamic environments due to traditional A* algorithms having major limitations including long computational time, generating paths with numerous sharp turns that are difficult for mobile robots to follow, and inability to handle dynamic obstacles in real-time. This research proposes a comprehensive solution comprising a modified A* algorithm combined with a fuzzy logic approach, including limiting neighborhood node expansion to only relevant directions, adding turn penalty functions to minimize sharp turns, path smoothing using G² Continuous Bezier Spiral (G2CBS), and implementing a fuzzy logic controller for dynamic obstacle avoidance. Test results demonstrate significant performance improvements with the modified A* successfully reducing computational time by up to 50% (from 13.7065s to 6.8971s), reducing node expansion by up to 21% (from 70,255 to 55,443 nodes), reducing the number of turns by up to 86.30% (from 73 to 10 turns), and improving path tracking accuracy with RMSE decreasing by up to 82.63% (from 0.0182 to 0.0033) when using a pure pursuit controller. Meanwhile, for the dynamic obstacle avoidance system, the mobile robot successfully avoided collisions at safe distances of 8.83cm (obstacle 1) and 7.33cm (obstacle 2) while achieving an RMSE of 0.929, demonstrating an optimal balance between safety and path tracking accuracy.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Mobile robot, Perencanaan Jalur, Penghindaran Rintangan, A* Modifikasi, Fuzzy Logic,Mobile robot, Path Planning, Obstacle Avoidance, Modified A, Fuzzy Logic
Subjects:	T Technology > TJ Mechanical engineering and machinery > TJ211.415 Mobile robots
Divisions:	Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Electrical Engineering > 20201-(S1) Undergraduate Thesis
Depositing User:	Muhammad Ikhsan
Date Deposited:	23 Jul 2025 08:41
Last Modified:	23 Jul 2025 08:41
URI:	http://repository.its.ac.id/id/eprint/120913

Actions (login required)

View Item