Isnawati, Salafiyah (2018) Model Hibrida Singular Spectrum Analysis dan Automatic ARIMA untuk Peramalan Air Terjual di PDAM Giri Tirta Sari Kabupaten Wonogiri Jawa Tengah. Undergraduate thesis, Institut Teknologi Sepuluh Nopember.
Preview |
Text
1314100076-Undergraduate_Theses.pdf - Accepted Version Download (2MB) | Preview |
Abstract
Semakin hari kebutuhan akan air bersih semakin meningkat tetapi ketersediaan air terbatas. Pada penelitian ini, PDAM Giri Tirta Sari Kabupaten Wonogiri dipilih dalam studi kasus karena terjadi pemborosan sumber daya air sehingga akan dilakukan peramalan air terjual dengan mengaplikasikan model hibrida SSA-Automatic ARIMA. Sebagai studi kasus digunakan data air terjual per bulan pada periode Januari 2006 sampai Agustus 2017. Terdapat dua kajian yang dilakukan pada penelitian ini, yaitu kajian simulasi dan kajian terapan. Kajian simulasi menunjukkan bahwa metode SSA lebih baik dilakukan secara agregat daripada secara individu. SSA mendekomposisi data ke dalam pola trend, musiman, dan noise, tetapi tidak dapat mendekomposisi variasi kalender. Pada kajian terapan digunakan empat metode pembanding yaitu Automatic ARIMA, SSA-ARIMA, ARIMA, dan ARIMAX. Data air terjual mengandung pola trend, musiman, noise, dan variasi kalender. SSA tidak dapat mendekomposisi pola variasi kalender, tetapi pola variasi kalender dapat ditangkap dengan baik oleh metode ARIMAX, sehingga metode terbaik untuk peramalan air terjual adalah metode ARIMAX.
=================================================================================================================
The need of clean water is increasing day by day, however its availability is limited. The objective of this research is apply hybrid SSA-Automatic ARIMA model for forecasting water demand at PDAM Giri Tirta Sari because there is a squander of water resources. The data used in this case study is the monthly water demand start from Januari 2006 up to August 2017. There are two studies in this research, that is simulation and applied studies. The result of simulation study show that SSA method is better done on an aggregat basis than an individual basis. SSA decompose the data into trend pattern, seasonal, and noise but it can not decompose the calendar variations pattern. There are four methods that be applied and compared in this research, that is Automatic ARIMA, SSA-ARIMA, ARIMA, and ARIMAX. Data of water demand consists of trend, seasonal, noise, and calendar variations pattern. SSA method can not decompose the calendar variations pattern, but this pattern can be well detected by ARIMAX method, so the best method for the forecast of water demand is ARIMAX method.
| Item Type: | Thesis (Undergraduate) |
|---|---|
| Additional Information: | RSSt 519.535 Isn m-1 3100018075069 |
| Uncontrolled Keywords: | Air Terjual; Automatic ARIMA; PDAM Wonogiri; Peramalan; Singular Spectrum Analysis; Automatic ARIMA; Forecasting; PDAM Wonogiri; water demand |
| Subjects: | Q Science > Q Science (General) T Technology > TD Environmental technology. Sanitary engineering > TD233 Water consumption |
| Divisions: | Faculty of Mathematics and Science > Statistics > 49201-(S1) Undergraduate Thesis |
| Depositing User: | salafiyah isnawati |
| Date Deposited: | 09 Apr 2018 03:51 |
| Last Modified: | 21 Sep 2020 06:20 |
| URI: | http://repository.its.ac.id/id/eprint/50621 |
Actions (login required)
 |
View Item |