Padma, Adinda Krisantya (2026) Sistem Take Off Landing Pada UAV Berbasis Aruco untuk Pengisian Daya Nirkabel. Other thesis, Institut Teknologi Sepuluh Nopember.
![[thumbnail of 5022221156_Undergraduate_Thesis.pdf]](http://repository.its.ac.id/style/images/fileicons/text.png) |
Text
5022221156_Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (4MB) | Request a copy |
Abstract
Penelitian ini mengusulkan sistem pendaratan otonom untuk UAV (Unmanned Aerial Vehicle) guna mengatasi keterbatasan daya melalui integrasi teknologi Wifi Channel State Information (CSI) menggunakan mikrokontroler ESP32 dan deteksi visual Aruco Marker pada drone DJI Tello. Sistem ini dirancang menggunakan pendekatan dua tahap, yaitu estimasi posisi jarak menengah (3–5 meter) memanfaatkan analisis karakteristik sinyal Wifi CSI, serta akurasi posisi akhir jarak dekat (0,1–3 meter) menggunakan deteksi visual Aruco Marker melalui kamera drone. Seluruh implementasi sistem diprogram menggunakan bahasa Python. Pengolahan data metode CSI fingerprinting fase offline menggunakan algoritma Random Forest Classifier berhasil mencapai tingkat akurasi evaluasi model yang tinggi sebesar 92,27%. Selain itu, modifikasi landing pad dengan lapisan aluminium foil terbukti secara signifikan meningkatkan efisiensi pantulan amplitudo sinyal sehingga mempermudah identifikasi posisi koordinat drone. Hasil pengujian menunjukkan bahwa sistem Hybrid localization ini mampu mengurangi kegagalan deteksi akibat kendala visual seperti pencahayaan rendah atau occlusion, serta mewujudkan siklus pendaratan otonom yang lebih adaptif dan robust untuk aplikasi industri.
==============================================================================================================================
This research proposes an autonomous landing system for UAVs (Unmanned Aerial Vehicles) to overcome power limitations thru the integration of Wifi Channel State Information (CSI) technology using the ESP32 microcontroller and visual detection of Aruco Markers on the DJI Tello drone. The system is designed using a two-stage approach, namely medium-range position estimation (3–5 meters) utilizing the analysis of Wifi CSI signal characteristics, and close-range final position precision (0.1–2 meters) using visual detection of Aruco Markers thru the drone's camera. The entire system implementation is programd using the Python language. The offline phase data processing of the CSI fingerprinting method using the Random Forest Classifier algorithm successfully achieved a high model evaluation accuracy rate of 92.27%. Additionally, the modification of the landing pad with an aluminum foil layer significantly improved the efficiency of signal amplitudo reflection, making it easier to identify the drone's coordinate position. The test results show that this Hybrid localization system is capable of reducing detection failures due to visual constraints such as low lighting or occlusion, and achieving a more adaptive and robust autonomous landing cycle for industrial applications.
Actions (login required)
 |
View Item |