Prediksi Kualitas Udara Di Kota Surabaya Berdasarkan Temperatur Menggunakan Model Fungsi Transfer Dan Feed Forward Neural Network

Wijaya, Galuh Indra (2019) Prediksi Kualitas Udara Di Kota Surabaya Berdasarkan Temperatur Menggunakan Model Fungsi Transfer Dan Feed Forward Neural Network. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 06211540000128-Undergraduate_Thesis.pdf Restricted to Repository staff only Download (9MB) | Request a copy

Abstract

Kota Surabaya sebagai ibu kota provinsi dan di kelilingi wilayah industri memiliki jumlah lalu lintas kendaraan yang tinggi, dimana gas buang kendaraan menghasilkan polusi. Polusi mempengaruhi kualitas udara yang ada. Kota Surabaya memiliki alat yang dapat memantau kualitas udara di stasiun SUF. Penelitian ini bertujuan untuk memantau lima pencemar udara utama yaitu PM10, CO, SO2, NO2, dan O3 yang dipengaruhi oleh temperatur udara dan memprediksi kualitas udara selama satu hari kedepan. Metode yang digunakan untuk meramalkan adalah metode fungsi transfer dan FFNN. Data yang digunakan pada penelitian ini berasal dari Dinas Lingkungan Hidup (DLH) mulai 1 Januari 2018 hingga 20 Desember 2018 berupa lima pencemar udara PM10, CO, SO2, NO2, dan O3 serta temperatur. Metode fungsi transfer menghasilkan ramalan kurang baik karena belum mampu menangkap pola data sedangkan ramalan metode FFNN sudah mampu menangkap pola data namun terdapat shifting satu periode. Hasil perbandingan secara RMSE, RMSEP, sMAPE, dan sMAPEP menunjukkan FFNN lebih baik karena memiliki nilai kebaikan model yang lebih kecil. Hasil peramalan menunjukkan kualitas udara di Kota Surabaya ditentukan oleh pencemar PM10 di SUF 1 & SUF 7 dan CO di SUF 6, secara umum kualitas udara terprediksi kurang baik pada pagi hingga siang hari.
=================================================================================================================================
Surabaya as the capital province of East Java surrounded by industrial areas and has high traffic. The vehicles produces pollution that affects air quality. Surabaya has tools that can monitor air quality at SUF stations. This study aims to monitor five main air pollutants which are PM10, CO, SO2, NO2, and O3 that influenced by temperature and aims to predict air quality of Surabaya for the next day. Methods to predict the air quality are Transfer Function and FFNN. The data used in this study originated from the Department of Environment (DLH) start from 1 January 2018 to 20 December 2018 of five air pollutants PM10, CO, SO2, NO2, & O3 and temperature. The transfer function method presents not so good predictions because it hasn't been able to capture data patterns while the FFNN method predictions have been able to capture data patterns though there are one-period shifts. After comparing the RMSE, RMSEP, sMAPE, and sMAPEP show that FFNN gives better prediction because it has less value in comparing values. Results show the air quality in Surabaya is determined by PM10 in SUF 1 & SUF 7 and CO in SUF 6, in general air quality is predicted to be moderate in the morning until dusk.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Air Polution, ARIMA, FFNN, Transfer Function
Subjects:	Q Science > QA Mathematics > QA276 Mathematical statistics. Time-series analysis. Failure time data analysis. Survival analysis (Biometry) Q Science > QA Mathematics > QA76.6 Computer programming. Q Science > QA Mathematics > QA76.87 Neural networks (Computer Science) T Technology > TD Environmental technology. Sanitary engineering > TD883.5 Air--Pollution
Divisions:	Faculty of Mathematics, Computation, and Data Science > Statistics > 49201-(S1) Undergraduate Thesis
Depositing User:	Galuh Indra Wijaya
Date Deposited:	23 Jul 2026 03:17
Last Modified:	23 Jul 2026 03:17
URI:	http://repository.its.ac.id/id/eprint/66192

Actions (login required)

View Item