PROTOTYPE EARLY WARNING FLOOD SYSTEM USING RAINFALL PARAMETERSWATER COLOR AND SURFACE HEIGHTWATER WITH THE K-MEANS METHOD NODEMCU BASED

Fauzan, Muhammad Restu and Sitorus, Meyhart Bangkit (2024) PROTOTYPE EARLY WARNING FLOOD SYSTEM USING RAINFALL PARAMETERSWATER COLOR AND SURFACE HEIGHTWATER WITH THE K-MEANS METHOD NODEMCU BASED. Diploma thesis, ITPLN.

[thumbnail of SKRIPSI 202011237 M.RESTU FAUZAN.pdf] Text
SKRIPSI 202011237 M.RESTU FAUZAN.pdf
Restricted to Registered users only

Download (9MB)

Abstract

Prototipe Early Warning System banjir adalah prototipe yang dapat mengirimkan peringatan dini akan bencana banjir. Prototipe ini terhubung dengan platform untuk fungsi monitoring dan penyimpanan data. Platform yang digunakan adalah Node-Red dan Firebase Database. Data pengujian dilakukan sebanyak 3 kali yang diolah dengan persamaan Euclidean Distance, metode Elbow, dan algoritma K-Means. Pengumpulan data dilakukan dengan memvariasikan variabel-variabel sensor jarak, curah hujan, dan warna air agar mendapatkan data yang beragam. Dari data pengujian tersebut diolah dengan algoritma K-Means menggunakan aplikasi RapidMiner. Klustering yang dihasilkan oleh algoritma K-Means akan diolah kembali dengan metode Elbow untuk menentukan nilai K yang paling tepat untuk data pengujian. Dengan merata-ratakan hasil keluaran setiap kluster akan menghasilkan nilai parameter-parameter yang dapat digunakan untuk memicu suatu tindakan. Hasil pengujian prototipe Early Warning System Banjir menunjukkan indikasi yang baik. Respon alat saat memasuki nilai-nilai parameter potensi banjir dan banjir sesuai dengan apa yang diharapkan seperti mengaktifkan LED, menghidupkan pompa juga menghidupkan buzzer alarm peringatan. Kondisi banjir digambarkan dengan parameter warna nilai sebesar 80,96, curah hujan sebesar 649,6, dan jarak sebesar 4,93 cm. Sementara untuk kondisi potensi banjir memiliki nilai parameter nilai warna sebesar 60,5, curah hujan sebesar 373,3, dan jarak sebesar 8,5 cm. Alat prototipe Early Warning System banjir memerlukan perkembangan bagian performa sistem. Terdapat jeda kerja yang terjadi saat alat berjalan sehingga mengganggu kerja sensor-sensor untuk mengukur variabelvariabelnya.

The Early Warning System flood prototype is a prototype that can send early warnings of flood disasters. This prototype is connected to a platform for monitoring and data storage functions. The platforms used are Node-Red and Firebase Database. Test data was carried out 3 times which were processed using the Euclidean Distance equation, the Elbow method, and the K-Means algorithm. Data collection was carried out by varying the distance sensor variables, rainfall, and water color to obtain diverse data. The test data was processed using the K-Means algorithm using the RapidMiner application. The clustering produced by the K-Means algorithm will be reprocessed using the Elbow method to determine the most appropriate K value for the test data. By averaging the output results of each cluster, it will produce parameter values that can be used to trigger an action. The results of the Early Warning System Flood prototype test showed good indications. The response of the tool when entering the potential flood and flood parameter values was in accordance with what was expected, such as activating the LED, turning on the pump and also turning on the warning alarm buzzer. Flood conditions are described by color parameter values of 80.96, rainfall of 649.6, and distance of 4.93 cm. Meanwhile, for potential flood conditions, the color parameter values are 60.5, rainfall of 373.3, and distance of 8.5 cm. The prototype Early Warning System flood tool requires development of the system performance section. There is a work gap that occurs when the tool is running, which interferes with the work of the sensors to measure the variables.

Item Type: Thesis (Diploma)
Uncontrolled Keywords: Prototipe Early Warning System banjir, Firebase Database, Node-Red, Algoritma K-Means, Klustering Flood Early Warning System Prototype, Firebase Database, Node-Red, KMeans Algorithm, Clustering
Subjects: Skripsi
Bidang Keilmuan > Teknik Elektro
Divisions: Fakultas Ketenagalistrikan dan Energi Terbarukan > S1 Teknik Elektro
Depositing User: Sudarman
Date Deposited: 28 Oct 2025 07:25
Last Modified: 28 Oct 2025 07:25
URI: https://repository.itpln.ac.id/id/eprint/3051

Actions (login required)

View Item
View Item