Investigation on the Integration of Low-Cost NIR Spectrometers in Mill Flour Industries for Protein, Moisture and Ash Content Estimation.

The flour milling industry-a vital component of global food production-is undergoing a transformative phase driven by the integration of smart devices and advanced technologies. This transition promises improved efficiency, quality and sustainability in flour production. The accurate estimation of protein, moisture and ash content in wheat grains and flour is of paramount importance due to their direct impact on product quality and compliance with industry standards. This paper explores the application of Near-Infrared (NIR) spectroscopy as a non-destructive, efficient and cost-effective method for measuring the aforementioned essential parameters in wheat and flour by investigating the effectiveness of a low-cost handle NIR spectrometer. Furthermore, a novel approach using Fuzzy Cognitive Maps (FCMs) is proposed to estimate the protein, moisture and ash content in grain seeds and flour, marking the first known application of FCMs in this context. Our study includes an experimental setup that assesses different types of wheat seeds and flour samples and evaluates three NIR pre-processing techniques to enhance the parameter estimation accuracy. The results indicate that low-cost NIR equipment can contribute to the estimation of the studied parameters.

An Approach of Non-Linear Systems Through Fuzzy Control Based on Takagi-Sugeno Method.

Today, the advanced technology is a part of the everyday's life. As a result, most of the applications used require a more complex system in order to achieve a better performance. These systems have a mathematic background indicating the need of a better mathematical tool to increase the reliability of them. One of the most significant problems coming up against these systems is undoubtedly the non-linearity of the equations governing them. Herein, a linearization method is proposed and studied through intelligent control. The transformation of a non-linear system into a linear is based on fuzzy logic and more specifically on Takagi-Sugeno technique. Firstly, it is analyzed in a theoretical level followed by two examples. The fuzzy model was developed through Matlab program. Finally, the efficiency of the above method was investigated setting up various values for the under study variables and comparing the results of them with the "actual" ones. The square error method was used for a better evaluation indicating that this method is a useful technique except from the applications where the high accuracy is mandatory.