Perancangan Sistem Pendaratan Presisi Drone Pada Platform Darat Bergerak Berbasis Visi Komputer

Abadi, Dionisius Yose Deofili (2020) Perancangan Sistem Pendaratan Presisi Drone Pada Platform Darat Bergerak Berbasis Visi Komputer. Undergraduate thesis, Institut Teknologi Sepuluh Nopember.

[img] Text
07111640000109-Undergraduate_Thesis.pdf - Accepted Version
Restricted to Repository staff only

Download (3MB) | Request a copy

Abstract

Pendaratan presisi merupakan salah satu manuver dari drone yang didefinisikan sebagai sistem pendaratan dimana drone mampu mendarat tepat di atas area pendaratan dengan variasi deviasi terkecil yang tidak memiliki nilai kuantitatif. Sistem ini dapat digunakan pada area perairan, konflik ataupun geografis yang rapuh dimana drone memiliki potensi hancur ketika mendarat di area tersebut. Sistem ini juga mampu menggantikan peran GPS pada area dimana sinyal satelit GPS lemah. Pada penelitian yang telah dilakukan, berbagai metode dengan mengkombinasikan GPS dengan sensor lain telah dilakukan tetapi belum dilakukan uji coba pada area pendaratan bergerak. Pada penelitian ini, dikembangkan metode koreksi posisi dengan mengkombinasikan komputer visi dengan sensor jarak untuk melakukan pendaratan presisi pada platform darat bergerak. Dengan mengkombinasikan x dan y yang didapat dari pengolahan citra dengan memanfaatkan penghitungan titik berat dan koordinat z dari sensor jarak, maka akan didapatkan sinyal kontrol untuk mengontrol drone. Dari pengujian, didapatkan hasil bahwa drone mampu melakukan koreksi posisi dengan baik dengan deviasi minimal hingga 10 cm tetapi akan mendapatkan deviasi besar apabila drone kehilangan area pendaratan saat menurunkan ketinggian sehingga area pendaratan menjadi tidak terdeteksi oleh kamera. Dapat disimpulkan bahwa drone mampu melakukan pendaratan presisi pada platform darat bergerak dengan kecepatan area pendaratan sekitar 10 cm/detik dengan deviasi rata – rata sebesar 35.25 cm dengan deviasi minimal 10 cm dan deviasi maksimal 103.5 cm. ======================================================================================================================== Precision landing is one of drone maneuvers that is determined by a landing system which drone is able to land directly above the landing area with the smallest deviation that has no quantitative value. This system can be used in water areas, conflict area, or fragile geographic areas which dronehave the potential to be destroyed if it is land on such places. This system is also able to replace the role of GPS in areas where GPS satellite signals are weak. In the recent research, various methods by combining GPS with other sensors have been done but have not been tested in the mobile landing area. In this research, a position correction method was developed with combining computer vision and proximity sensor to perform precision landing on moving platform. By combining x and y form image processing by utilizing the calculation of center of gravity an z coordinate from proximity sensor, a control signal will be obtained to control the drone. From the test, the result are obtained that the drone is able to make a good position correction with minimum deviation up to 10 cm but will get a large large deviation if the drone lost the sight of landing area so that the landing area cannot be tracked by the camera. It can be concluded that the drone is able to make a precision landing on a moving platform with a landing area velocity about 10 cm/sec with an average deviation of 35.25 cm with a minimum deviation of 10 cm and a maximum deviation of 103.5 cm.

Item Type: Thesis (Undergraduate)
Uncontrolled Keywords: drone, platform darat bergerak, visi komputer drone, ground moving platform, computer vision
Subjects: U Military Science > UG1242 Drone aircraft--Control systems. (unmanned vehicle)
U Military Science > U Military Science (General) > UG Military Engineering > UG1242.D7 Unmanned aerial vehicles. Drone aircraft
Divisions: Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Electrical Engineering > 20201-(S1) Undergraduate Thesis
Depositing User: Dionisius Yose Deofili Abadi
Date Deposited: 06 Aug 2020 08:20
Last Modified: 06 Aug 2020 08:20
URI: https://repository.its.ac.id/id/eprint/77102

Actions (login required)

View Item View Item