Fuzzy Mamdani based user-friendly interface for food preservatives determination

Abstract

Several food processing manufacturers have added extra preservatives to extend the shelf life and maintaining the quality of their products. Although those preservatives may preserve the food longer, the implication when consuming them in a large amount will affect our health. They can cause headaches, palpitations, allergies and even cancer. Therefore, it is necessary to establish an alternative method, which will provide the information on the content of preservatives to consumers and auditors. In addition, it will prevent the manufacturers in adding additional preservatives without the knowledge of the authorities. Thus, this study proposed a fuzzy based algorithm to determine the percentage of preservatives in processed fruits. The fuzzy logic framework is developed based on Mamdani inference system and is compared to Takagi-Sugeno-Kang (TSK) inference. Once the algorithm has been completed, a Graphical User Interface (GUI) will be developed based on the fuzzy framework aiming to create a user-friendly interface for determining the percentage of food preservatives. It is observed that fuzzy logic has successfully determined the percentage of preservatives namely sulphur dioxide, benzoic acid and sorbic acid in processed fruits. From the result, the average total absolute error of Mamdani system is about 2.31%, which is smaller than Sugeno system, which is about 19.51%. Thus, Mamdani based inference system showed a better performance than Takagi-Sugeno-Kang (TSK). In addition, it is well-suited to human related problems especially in determining the percentage of food preservatives. Furthermore, the Mamdani based GUI has been successful developed to determine the percentage of food preservatives with simple formulation, fast, user-friendly and accurately align with the standards as outlined in Malaysia Food Act 1983.