ARTICLE
TITLE

Forecasting Produksi Energi PLTS 1 kWp Menggunakan Mesin Pembelajaran Dengan Algoritma Support Vector Machine

SUMMARY

AbstrakIsu krisis energi menuntut orang mencari sumber energi alternatif, PLTS menjadi pilihan yang menjanjikan untuk menjawab tantangan krisis energi tersebut. Namun PLTS tergantung oleh kondisi cuaca, sangat sulit memperkirakan berapa produksi energi pada suatu system PLTS. Penelitian peramalan produksi energi 1kWp menggunakan mesin pembelajaran dan support vector machine (SVM) telah dilakukan dan dibandingkan dengan model multiple linear regression (MLR), model peramalan dengan pendekatan deret waktu, data training periode Januari – Desember 2021 dan data tes periode Januari – Mei 2022. Konstruksi ramalannya adalah hasil produksi tiga hari kebelakang meramalkan produksi energi hari berikutnya. Hasil evaluasi MAPE pada data training SVM dan MLR adalah 19.79% dan 23.96%, sedangkan pada testing 21,79% dan 20.45%. Hasil peramalan harian diakumulasi perbulan dan dievaluasi, hasilnya MAPE 4.13% dan 5.56% masing-masing untuk SVM dan MLR. Kedua model SVM dan MLR layak dikembangkan lebih lanjut pada forecasting PLTS 1 kWp berdasarkan data deret waktu.AbstractThe issue of the energy crisis requires people to look for alternative energy sources, SPP is a promising choice to answer the challenges. However SPP depends on weather conditions, it is very difficult to estimate energy production. Research on forecasting 1kWp energy production using machine learning and SVM has been carried out and compared with MLR, forecasting models with a time series approach, training data for the period January – December 2021 and test data for the period January – May 2022. MAPE evaluation results on SVM and MLR training data were 19.79% and 23.96%, while testing was 21.79% and 20.45%. Daily forecasting results are accumulated monthly and evaluated, the results are MAPE 4.13% and 5.56% for SVM and MLR, respectively. Both SVM and MLR models deserve to be further developed in forecasting PLTS 1 kWp based on time series data.

 Articles related

Mukhammad Jainudin, Nuriyanto    

Xxx is one of the companies engaged in construction that produces paving stones. The obstacle experienced by this company is that demand is sometimes soaring and sometimes soaring is caused by a lack of careful planning and maximum profits derived by the... see more


Erik Wijayanti, Khafidz Rosydi    

The problem being faced is the product demand that is not constant, so that production planning becomes difficult. The result is often the difference between the amount of demand and product inventory. The company must find an effective and efficient pro... see more


Yufri Alrahman,Kamil Mustafa,Yuana Delvika    

PT. CJ Feed Medan is a engaged in the animal feed industry. The purpose of this study is to determine the amount of GM-1C feed production from the forecasting of the number of sales of animal feed products of the previous type. And to plan raw material i... see more


Erni Yusnita,Juarni Juarni    

CV. Kotama Shoes is a trading company engaged in the production of leather shoes. The company is located on Jalan A.R. Judge No.112c / 206c. In running its business, CV. Kotama Shoes has several problems or constraints in production planning. Erratic flu... see more