Skripsi

SISTEM PREDIKSI BANJIR MENGGUNAKAN SENSOR TEKANAN DAN WATERFLOW SENSOR DENGAN METODE DECISION TREE

---

Most current flood prediction studies rely on secondary data with limited temporal resolution and Deep Learning models with high computational demands. This research addresses this gap by designing an early warning system that is computationally efficient, utilizing direct physical sensor data within a controlled simulation. The objectives are to measure water flow velocity (m/s) using an optocoupler-based waterflow sensor, estimate water level (m) using a BMP280 pressure sensor, and build a flood classification model using the Decision Tree algorithm. The architecture integrates a NodeMCU ESP32 with a PC via a hybrid serial USB and TCP/IP-based local Wi-Fi network. Sensor data is processed in real-time by a Python script into "AMAN" (SAFE), "WASPADA" (ALERT), and "BANJIR" (FLOOD) statuses, then recorded into a .csv file. Results show optimal hardware performance, with average errors of 4.70% and 6.70% for the optocoupler, and 3.90% for the BMP280. The Decision Tree model simulated in Google Colab achieved a high accuracy of 97%. This system is proven to be interpretable and efficient for local hydrological monitoring applications.

---

Availability

Detail Information

Series Title

-

Call Number

T2005362026

Publisher

Indralaya : Prodi Sistem Komputer, Fakultas Ilmu Komputer Universitas Sriwijaya., 2026

Collation

xv, 100 hlm.; ilus.; tab.; 29 cm.

Language

Indonesia

ISBN/ISSN

-

Classification

005.740 7

Content Type

Text

Media Type

-

Carrier Type

-

Edition

-

Subject(s)

Prodi Sistem Komputer
Sistem Prediksi Banjir

Specific Detail Info

-

Statement of Responsibility

MI

Other version/related

File Attachment

Comments

---

You must be logged in to post a comment