Perencanaan Jalur dan Penghindaran Rintangan Dinamis Berbasis Improved Informed-RRT* dan Improved DWA untuk Robot Nonholonomik

Ghaly, Muhammad Fawwazul (2026) Perencanaan Jalur dan Penghindaran Rintangan Dinamis Berbasis Improved Informed-RRT* dan Improved DWA untuk Robot Nonholonomik. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5022221011-Undergraduate_Thesis.pdf Restricted to Repository staff only Download (7MB) | Request a copy

Abstract

Navigasi nonholonomic mobile robot di lingkungan dinamis memerlukan integrasi antara perencanaan lintasan global dan penghindaran rintangan lokal. Algoritma Informed RRT* (IRRT*) sebagai perencana global masih menghasilkan belokan tajam dan waktu komputasi tinggi serta belum menangani rintangan dinamis, sedangkan Dynamic Window Approach (DWA) sebagai perencana lokal tidak memiliki perencanaan global. Penelitian ini mengusulkan integrasi Improved IRRT* dan Improved DWA untuk mengatasi keterbatasan tersebut. Improved IRRT* dirancang dengan goal bias, APF steering, adaptive step size, greedy strategy, dan penghalusan kurva Bezier G2CBS derajat tujuh untuk menghasilkan lintasan efisien dan halus sesuai kendala kinematik. Improved DWA dikembangkan dengan fungsi evaluasi yang memisahkan rintangan statis dan dinamis, serta komponen kedekatan lintasan global untuk menghindari tabrakan. Sistem diuji pada model lingkungan ruangan B104 menggunakan ROS2 Jazzy melalui Gazebo dan RViz dengan skenario rintangan menyilang dan berhadapan. Hasil pengujian menunjukkan Improved IRRT* mengurangi jumlah node yang diekspansi hingga 56.7% dan waktu komputasi hingga 20.5%, serta mengeliminasi seluruh belokan tajam dengan panjang lintasan tetap efisien. Improved DWA mampu mencapai tujuan tanpa tabrakan dengan pelacakan lintasan lebih baik, ditunjukkan oleh peningkatan efisiensi hingga 4.5%,penurunan deviasi panjang lintasan hingga 85.5%, dan penurunan CTE RMSE hingga 54.8% dibandingkan DWA konvensional. Pendekatan integratif Improved IRRT*-Improved DWA terbukti menghasilkan navigasi nonholonomic mobile robot yang efisien, aman, dan adaptif pada lingkungan kompleks dan dinamis.
==================================================================================================================================
Navigation of nonholonomic mobile robots in dynamic environments requires the integration of global path planning and local obstacle avoidance. The Informed RRT* (IRRT*) algorithm as a global planner still produces sharp turns and high computation time and does not yet handle dynamic obstacles, whereas the Dynamic Window Approach (DWA) as a local planner lacks a global planning mechanism. This study proposes the integration of Improved IRRT* and Improved DWA to overcome these limitations. Improved IRRT* is designed with goal bias, APF steering, adaptive step size, greedy strategy, and seventh-degree G2CBS Bézier curve smoothing to generate efficient and smooth paths that satisfy kinematic constraints. Improved DWA is developed with an evaluation function that separates static and dynamic obstacles, along with a global-path-proximity component to avoid collisions in real time. The system is tested on a model of the B104 room environment using ROS2 Jazzy through Gazebo and RViz with crossing and head-on obstacle scenarios. The test results show that Improved IRRT* reduces the number of expanded nodes by up to 56.7% and computation time by up to 20.5%, while eliminating all sharp turns with the path length remaining efficient. Improved DWA is able to reach the goal without collision and with better path tracking, indicated by an efficiency improvement of up to 4.5%, a path-length deviation reduction of up to 85.5%, and a CTE RMSE reduction of up to 54.8% compared to conventional DWA. Overall, the integrative Improved IRRT*-Improved DWA approach is proven to produce efficient, safe, and adaptive navigation of nonholonomic mobile robots in complex and dynamic environments.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Improved DWA, Improved IRRT*, Perencanaan Lintasan, Penghindaran Rintangan, Robot Nonholonomic, Improved DWA, Improved IRRT*, Nonholonomic mobile robot, Obstacle Avoidance, Path Planning.
Subjects:	T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK3070 Automatic control
Divisions:	Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Electrical Engineering > 20201-(S1) Undergraduate Thesis
Depositing User:	Muhammad Fawwazul Ghaly
Date Deposited:	23 Jul 2026 09:10
Last Modified:	23 Jul 2026 09:10
URI:	http://repository.its.ac.id/id/eprint/136482

Actions (login required)

View Item