Avaliação de método para determinação de açúcares redutores totais em cana de açucar por espectrscopia de infravermelho próximo (NIR)

Abstract

This study aimed to evaluate the technical and economic feasibility of using Near-Infrared Spectroscopy (NIR) for the determination of total reducing sugars (TRS) in sugarcane, as an alternative to the traditional Consecana method used in sugar cane. The conventional method involves several analytical steps, such as sampling, clarification, and the use of specific equipment, making it time-consuming and resource-intensive. In contrast, NIR is a faster, nondestructive method with the potential to replace the traditional procedure. To validate the equivalence of both methods, statistical tests, including the paired t-test and Machine Learning algorithms using artificial neural networks, were applied to compare data from both techniques. The results demonstrated that NIR produces data statistically comparable to the Consecana method, with significant reductions in operational costs and processing time. The study concluded that implementing the NIR method in sugarcane processing plants is both technically and economically viable, offering a more efficient and cost-effective alternative for TRS determination, with potential benefits for the sustainability and profitability of the sugar-energy sector.