Fuzzy Logic Approach for Banana Grading

Abstract

Banana fruit is most cultivated fruit in the world. Banana is most demandable because of its high nutrient values, easily availability and lower cost.  In any stage banana is useful for multiple products hence it is necessary to grade them properly. In this paper Fuzzy Logic is applied to grade the bananas. The chemical characteristics such as Total Soluble Sugar (TSS) and pH value were measured using handheld refractometer and pH meter. Triangular membership function of Fuzzy Logic Interface is used for Input and Output variables. For decision making Mamdani Interface was used and for defuzzification Centroid method was used. Linear Regression analysis was done and obtained the value 0.85 for adjusted R square. The ripening of banana increases with increase in TSS and pH value. Based on this banana were graded as unripe, partially ripe, fully ripe, and over ripe.   Designed fuzzy interface system showed 98.3% correct grading. The designed system is useful to grade the banana and use them to produce different banana biproducts based on its ripening stage.