Gupita, Sari Angelina Nurma (2017) Prediksi Kadar Polutan Menggunakan Adaptive Neuro Fuzzy Inference System (ANFIS) untuk Pemantauan Kualitas Udara di Kota Surabaya. Undergraduate thesis, Institut Teknologi Sepuluh Nopember.
Preview |
Text
2413100016-Undergraduate_Theses.pdf - Published Version Download (2MB) | Preview |
Abstract
Pencemaran udara merupakan salah satu permasalahan yang dijumpai di kota besar tak terkecuali Surabaya. Sumber pencemaran udara dapat berasal dari berbagai kegiatan industri, transportasi, perkantoran, dan perumahan. Kualitas udara di Surabaya dipantau melalui air quality monitoring system (AQMS) yang dilakukan oleh Badan Lingkungan Hidup (BLH). Penelitian ini mengusulkan sebuah model prediktor kualitas udara di Kota Surabaya. Variabel yang diprediksi adalah O3 (Ozon) dengan metode Adaptive Neuro-Fuzzy Inference System (ANFIS).
Terdapat sembilan variabel yang digunakan untuk masukan pada sistem prediktor, yaitu: konsentrasi NO2, SO2, O3, NO, arah angin, global radiasi, kecepatan angin, temperature, dan kelembaban udara kota Surabaya. Perancangan model dilakukan melalui 2 tahap, yaitu pelatihan terhadap 9 model dan validasi. Pelatihan dan validasi dilakukan dengan menggunakan data yang diambil sejak 1 Januari 2012 sampai 31 Desember 2016. Data diolah untuk menghilangkan missing value dan outliers. Data yang digunakan untuk pelatihan sebanyak 504 data dan untuk validasi sebanyak 216 data. Parameter pada prediktor ditentukan oleh nilai range of influence. Akurasi tertinggi model diperoleh saat menggunakan 5 variabel masukan yakni kadar O3 , NO2, SO2, kelembaban, dan temperatur. Model ini menghasilkan error terendah yakni sebesar 11.62% atau nilai akurasi sebesar 88.38% pada nilai range of influence 0.7.
=================================================================
Air pollution is one of common problem found in big city like Surabaya. Pollution sources come from industrial activities, transportation, office, and housing. Air quality in Surabaya is monitored using air quality monitoring system (AQMS) held by Badan Lingkungan Hidup (BLH). This research suggest an air quality predictor model. Predicted variable is O3 (Ozone) using Neuro-Fuzzy Inference System (ANFIS) method.
There are 9 variables used as input variables in this predictor system, they are: NO2, SO2, O3, NO, wind direction, global radiation, wind velocity, temperature, and humidity. Model design consist of 2 steps, training and validating to 9 models. Training and validating using data from 1st January 2012-31st December 2016. The data was processed to delete missing value and outliers. 504 remains data used as training data, and 216 data are used as validating data. Predictor parameters are determined by range of influence value. The highest accuracy reached when 5 variables were used. They are: O3 , NO2, SO2, humidity and temperature. This model produce the lowest error value 11.62% or 88.38% as the accuracy value at range of influence value 0.7.
| Item Type: | Thesis (Undergraduate) |
|---|---|
| Additional Information: | RSF 629.89 Gup p |
| Uncontrolled Keywords: | akurasi, ANFIS, Ozon, polutan, prediksi, range of influence |
| Subjects: | T Technology > TL Motor vehicles. Aeronautics. Astronautics > TL725.3 Traffic Control |
| Divisions: | Faculty of Industrial Technology > Physics Engineering > 30201-(S1) Undergraduate Thesis |
| Depositing User: | Sari Angelina Nurma Gupita |
| Date Deposited: | 12 Dec 2017 07:04 |
| Last Modified: | 05 Mar 2019 06:49 |
| URI: | http://repository.its.ac.id/id/eprint/47538 |
Actions (login required)
 |
View Item |