Search Repository STTMIGAS Balikpapan

Saat ini, bahan bakar minyak (BBM) masih sangat dibutuhkan oleh masyarakat luas
tak terkecuali di Indonesia. PT Pertamina, sebagai BUMN yang ditugaskan untuk
menyediakan kebutuhan akan BBM di Indonesia telah mengenalkan bermacam-macam jenis
BBM, namun beberapa yang paling sering kita dengar diantaranya Premium, Pertamax , Solar
dan yang terbaru di tahun 2016 adalah Pertalite. Solar adalah salah satu jenis BBM yang
disediakan oleh PT. Pertamina (Persero) untuk kebutuhan nasional. PT. Pertamina (Persero)
adalah satu-satunya Badan Usaha Milik Negara (BUMN) yang memproduksi BBM solar.
Maka dari itu dilakukan peramalan untuk memprediksikan permintaan solar 5 periode
kedepan dari tanggal 1 sampai 5 mei dengan melakukan pemilihan metode peramalan yang
tepat antara metode Linear Regression dengan metode spesifik Simple linear Regression dan
Non-Linear Regression dengan metode spesifik Exponential Regression, Logarithmic
Regression, Polynomial Regression, dan Power Regression.
For the application of this method can be done to predict demand forecasting future
periods 5 using WinQSB software , and to make the selection method is precisely seen from
the types of forecasting error rate (errors forecasting) to find the value of MAD , MSE and
MAPE smallest . After forecasting error , the method of Non - Linear Regression with specific
methods , namely Power Regression chosen because it has the value of MAD and MAPE
smallest compared to other methods. Nilai dari error forecasting Power Regression yaitu nilai
MAD = 66.759,67; MSE = 7.642.974.914,77 dan MAPE = 38,14%. Hasil peramalan Power
Regreesion untuk 5 periode mei kedepan pada tanggal 1-5 mei adalah 198.445,07 bbl;
198.391,62 bbl; 198.338,62 bbl; 198.286,05 bbl; 198.233,92 bbl.

Currently , fuel oil ( BBM ) is still very much needed by the wider community
including in Indonesia . PT Pertamina , as a state that is assigned to will supply the fuel in
Indonesia has introduced various types of fuel , but some of the most frequently heard among
Premium, PERTAMAX, Solar and most recently in 2016 is Pertalite. Solar is one type of fuel
that is supplied by PT . Pertamina ( Persero ) for national needs . PT . Pertamina ( Persero ) is
the only State-Owned Enterprises ( SOEs ) that produce diesel fuel .
Therefore do forecasting to predict demand for solar 5 periods ahead from 1 to 5 May
to conduct elections correct forecasting method among methods Linear Regression with
specific methods Simple Linear Regression and Non - Linear Regression with specific
methods Exponential Regression , logarithmic Regression , Polynomial Regression , and
Power Regression.
For the application of this method can dilakukkan forecasters to predict future
demand for 5 periods using WinQSB software , and to make the selection method is precisely
seen from the types of forecasting error rate (errors forecasting) to find the value of MAD ,
MSE and MAPE smallest . After forecasting error , the method of Non - Linear Regression
with specific methods , namely Power Regression chosen because it has the value of MAD
and MAPE smallest compared to other methods. Values of error Power Regression forecasting
a value MAD = 66759.67 ; MSE = 7,642,974,914.77 and MAPE = 38.14 % . Power
Regreesion forecasting results for the period May 5 ahead on 1-5 May was 198,445.07 bbl ;
198,391.62 bbl ; 198,338.62 bbl ; 198,286.05 bbl ; 198,233.92 bbl .