Metode Penghindaran Rintangan Pada Robot Mobil Beroda Berdasarkan Headed Social Force Model Dengan Pengendali Kecepatan Menggunakan Hybrid Velocity Obstacles

Fuad, Muhammad (2021) Metode Penghindaran Rintangan Pada Robot Mobil Beroda Berdasarkan Headed Social Force Model Dengan Pengendali Kecepatan Menggunakan Hybrid Velocity Obstacles. Doctoral thesis, Institut Teknologi Sepuluh Nopember.

Text 07111660010013-Dissertation.pdf - Accepted Version Restricted to Repository staff only until 1 October 2023. Download (2MB) | Request a copy

Abstract

Penelitian pada bidang penghindaran rintangan secara umum berasumsi bahwa objek rintangan memiliki kecepatan konstan dan hanya menyentuh sisi fisik manusia. Pada kenyataannya, individu atau kelompok manusia bergerak dengan kecepatan tidak konstan dan menggunakan lintasan nonlinear. Penelitian
ini bertujuan untuk menghasilkan metode penghindaran yang dapat menentukan kecepatan linear dan kecepatan sudut bebas benturan terhadap objek statis maupun rintangan bergerak nonlinear guna menjamin keselamatan fisik dan kenyamanan psikologis bagi manusia di sekitar robot.
Penelitian ini memodifikasi Velocity Obstacles (VO) menjadi Modified Steering VO (MSVO) dengan menambahkan perhitungan kecepatan sudut untuk melakukan steering berdasarkan informasi jarak dari sensor LIDAR guna menghindari rintangan statis. Untuk mengantisipasi gerakan objek rintangan multi
modalitas dengan lintasan nonlinear, penelitian ini mengusulkan Hybrid Velocity Obstacles (HVO). Headed Social Force Model (HSFM) dimodifikasi dengan mengganti desired velocity yang konstan dengan kecepatan linear bebas benturan
dari HVO untuk menghasilkan gaya tarik target. Gaya tolak terhadap rintangan juga memperhitungkan interaksi manusia-objek serta kelompok pedestrian selain individu pedestrian dan rintangan statis. Threat Level Index (TLI) diusulkan
sebagai indikator dalam menilai tingkat keselamatan fisik dan kenyamanan psikologis manusia di sekitar robot.
Untuk mengevaluasi metode usulan, beberapa skenario simulasi telah dikerjakan dengan mengimplementasikan MSVO, HVO, dan HSFM termodifikasi berbasis HVO pada robot mobil beroda dua berpenggerak differential-steering yang beroperasi di lingkungan manusia dalam ruangan. Hasil simulasi menunjukkan bahwa MSVO mampu menghindari benturan terhadap objek statis.
HVO dapat menjaga keselamatan fisik objek bergerak multi modalitas dengan rerata nilai TLI 0,17. HSFM termodifikasi berbasis HVO dapat memelihara kenyamanan psikologis kelompok pedestrian yang dalam kondisi terganggu dengan menghasilkan nilai TLI 0,14. Nilai TLI ini menyatakan bahwa robot mampu memelihara keselamatan dan kenyamanan manusia di sekitarnya.
=====================================================================================================
Research in the field of obstacle avoidance generally assumes that obstacle objects have a constant speed and only consider the physical side of humans. In reality, individuals or groups of humans move at non-constant speeds and use nonlinear trajectories. This study aims to produce an avoidance method that can determine collision-free linear velocity and angular velocity against static objects and nonlinear moving obstacles in order to ensure physical safety and psychological comfort for humans around the robot.
This study modifies Velocity Obstacles (VO) to Modified Steering VO (MSVO) by adding angular velocity calculations to perform steering based on distance information from the LIDAR sensor to avoid static obstacles. To anticipate the movement of multimodality obstacle objects with nonlinear
trajectories, this study proposes Hybrid Velocity Obstacles (HVO). The Headed Social Force Model (HSFM) was modified by replacing the desired velocity constant with the collision-free linear velocity of the HVO to produce the target pull force. The repulsive force against obstacles also takes into account human-object interactions and pedestrian groups in addition to individual pedestrians and static obstacles. Threat Level Index (TLI) is proposed as an indicator in assessing the level of physical safety and psychological comfort of humans around robots.
To evaluate the proposed method, several simulation scenarios have been carried out by implementing MSVO, HVO, and modified HSFM based on HVO on a two-wheeled mobile robot with differential-steering drive operating in an indoor human environment. The simulation results show that MSVO is able to avoid collisions with static objects. HVO can maintain the physical safety of multimodal moving objects with an average TLI value of 0.17. The modified HSFM based on HVO can maintain the psychological comfort of pedestrian
groups who are in a disturbed condition by producing a TLI value of 0.14. This TLI value states that the robot is able to maintain the safety and comfort of humans around it.

Item Type:	Thesis (Doctoral)
Uncontrolled Keywords:	Headed Social Force Model, Hybrid Velocity Obstacles, Lintasan Nonlinear, Modified Steering Velocity Obstacles, Penghindaran Rintangan, Robot Mobil Beroda differential-steering.Headed Social Force Model, Hybrid Velocity Obstacles, Nonlinear Track, Modified Steering Velocity Obstacles, Obstacle Avoidance, Differential-steering Wheeled Mobile Robot.
Subjects:	T Technology > TJ Mechanical engineering and machinery > TJ211.415 Mobile robots
Divisions:	Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Electrical Engineering > 20001-(S3) PhD Thesis
Depositing User:	Muhammad Fuad
Date Deposited:	18 Aug 2021 13:55
Last Modified:	18 Aug 2021 13:55
URI:	http://repository.its.ac.id/id/eprint/87572

Actions (login required)

View Item