Pengembangan Model Prediksi Kadar Air, Glukomanan, dan Kalsium Oksalat pada Tepung Porang secara Non-Destruktif Menggunakan Metode Spektroskopi Near Infrared

KHOIRUNNISAA, Hanim Zuhrotul Amanah, S.T.P., M.P., Ph.D.; Dr. Sri Rahayoe, S.T.P., M.P.

2024 | Skripsi | TEKNIK PERTANIAN

Tepung porang merupakan salah satu produk turunan umbi porang yang sedang digencarkan sebagai komoditas ekspor. Kualitas tepung porang ditentukan dari kadar air, glukomanan, dan kalsium oksalat. Pengujian kualitas tersebut umumya dilakukan dengan metode destruktif sehingga membutuhkan waktu yang lama, tidak realtime, dan hanya representatif dari sekelompok sampel. Metode nondestruktif spektroskopi Near Infrared dengan Panjang gelombang 700—2500 nm cocok dilakukan untuk pengujian kualitas tepung porang karena memerlukan waktu yang singkat dan hasilnya realtime. Penelitian ini dilakukan untuk membangun model prediksi kandungan glukomanan, kalsium oksalat, dan air pada tepung porang dengan melakukan preprocessing dan analisis menggunakan metode Partial Least Square Regression (PLSR). Sampel tepung porang yang digunakan untuk analisis kadar air yaitu sampel yang disimpan pada tempat lembap dengan variasi waktu penyimpanan selama 20 hari, sedangkan sampel untuk analisis kadar glukomanan dan kalsium oksalat yaitu dari 54 jenis tepung porang. Hasil penelitian model prediksi kadar air diperoleh nilai R2 0,983, kadar glukomanan diperoleh nilai R2 0,861, dan kadar kalsium oksalat diperoleh nilai R2 0,789. Dari hasil R2 tersebut dapat disimpulkan bahwa metode preprocessing yang digunakan dapat memperbaiki model, dibuktikan dari perhitungan nilai Ftest > Ftabel. 

Porang flour is one of the derivative products of porang tubers that is being promoted as an export commodity. The quality of porang flour is determined by the water content, glucomannan, and calcium oxalate. The quality testing is generally carried out using destructive methods so that it takes a long time, is not real-time, and is only representative of a group of samples. The non-destructive method of Near Infrared spectroscopy with a wavelength of 700-2500 nm is suitable for testing the quality of porang flour because it takes a short time and the results are real-time. This study was conducted to build a prediction model for the content of glucomannan, calcium oxalate, and water in porang flour by preprocessing and analyzing using the Partial Least Square Regression (PLSR) method. The porang flour samples used for water content analysis were samples stored in a humid place with a storage time variation of 20 days, while the samples for glucomannan and calcium oxalate content analysis were from 54 types of porang flour. The results of the water content prediction model obtained an R2 value of 0.983, glucomannan content obtained an R2 value of 0.861, and calcium oxalate content obtained an R2 value of 0.789. From the R2 results, it can be concluded that the preprocessing method used can improve the model, as evidenced by the calculation of the Ftest value > Ftable.

Kata Kunci : Tepung porang, Spektroskopi NIR, PLSR, Preprocessing