Multi‐Objective Optimization of a Continuous Amaranth Grain Puffing Process Using Taguchi– <scp>GRA</scp> Approach

ABSTRACT Amaranth grains can be thermally processed by several methods, such as oven puffing, gun puffing, extrusion puffing, oil puffing, and pan popping, but their small size, puffing characteristics, and low volume expansion ratio limit the effectiveness of these conventional techniques. In this study, Amaranth grains are puffed using a specially designed Special Purpose Machine (SPM) with a heating plate, developed to overcome these limitations and to enable controlled, continuous processing. The key process variables considered are the heating plate temperature, plate angle, nozzle hole size, and plate material, which strongly influence the popping yield (PY), volume expansion ratio (VER), and sensory score (SS) of the grains. These quality indicators are evaluated to understand the influence of process parameters and to establish a basis for systematic optimization. The primary objective of this work is to identify and optimize the most influential process parameters governing the puffing quality of Amaranth grains. The Taguchi design of experiments framework is employed to study the effect of multiple process variables on the grain quality, using an L16 orthogonal array to minimize the number of experiments while capturing the main effects. Significant factors are identified through signal‐to‐noise ratio analysis and analysis of variance (ANOVA), carried out using Minitab 17 software. Gray Relational Analysis (GRA) is used to perform multiobjective optimization by combining PY, VER, and SS into a single performance index. Experimental analysis and optimization indicate that the design of the process parameters can be significantly improved by jointly considering PY, VER, and SS. An L16 orthogonal array is implemented, and the initial parameter combination P4Q2R3S2 yields a gray relational grade (GRG) of 0.778. By adopting the optimized combination P3Q4R3S2, the GRG increases to 0.790, corresponding to a 1.89% improvement in overall performance. These results confirm that the design factors have been effectively optimized using Gray Relational Analysis, leading to improved machine settings for the Lahi machine and enhanced puffing quality of Amaranth grains. This work applies Gray Relational Analysis in conjunction with the Taguchi method for the multiobjective optimization of Amaranth grain puffing, demonstrating a quantifiable 1.89% improvement in quality indicators while using a specially designed SPM‐based heating system.