Modelling and Optimizing the Integrity of an Automated Vegetable Leaf Packaging Machine

Abstract

This study emphasized the need for postharvest technology in Nigeria's vegetable production to reduce postharvest losses ranging from 5% to 50%, focusing on enhancing processes of automated packaging unit of vegetable processing plant through the use of artificial neural networks (ANN). The experiment was conducted on a vegetable leaf processing plant with the objective of improving the reliability and performance of the automated packaging unit. Operating parameters such as moisture contents, leave particle size, time taken, throughput capacity, and specific mechanical energy consumption were varied to determine the optimum condition for each parameter. Statistical analysis was performed using R software. The appropriate model was chosen based on selection of the highest coefficient of prediction where the additional terms are significant and the model was not aliased, insignificant lack of fit and the maximization of the “Adjusted R2 value” and the “Predicted R2 value.” An optimum packaging condition was obtained at 15% moisture content, and 104.4 particle sizes which gave an optimum packaging time of 0.02 h, optimum packaging capacity of 57.31 kg/h, optimum SMEC value of 0.008 kw/h/kg, optimum repeatability value of 0.128 kg, optimum linearity value of 4.713 cm, optimum accuracy value of 5.2 cm (±0.45). The performance of the ANN model was evaluated using various measures such as mean squared error (MSE), the coefficient of determination (R2), mean absolute error (MAE), and the adjusted R‐squared (Adj. R2) for packaging machine. The results of this study suggest that ANN can be used to effectively optimize packaging units of the vegetable leaf processing plant.