An artificial neural network for optimizing safety and quality in thermal food processing

Abstract

Presents a backpropagation artificial neural network for optimizing food safety and quality in thermal processing applications. Five inputs (can size, initial temperature, thermal diffusivity, sensitivity indicator of microorganism, and sensitivity indicator of quality) are used as inputs to the network. The network computes the optimal control parameters (sterilization temperature, process time) and quality degradation of the food. This study is based on a wide range of microorganisms involved in foods.