Designing the Thermal Process for Sterilizing Canned Fish in the Autoclave: Modeling Temperatures with Artificial Neural Networks, Thermal Penetration Characteristics and Predicting Shelf Life

The current study aimed to sterilize canned fish cans using an autoclave, as well as addressing the thermal penetration characteristics and predicting shelf life. Thermal penetration characteristics were study, along with the calculation of their thermal process, and the application of artificial intelligence (ANN) was used to predict the sterilizer and cold spot temperature, as well as the lethal rate. The cold spot in the cans and the lethal rate during storage were investigated. The autoclave was used to sterilize the fish at two temperatures, 110 and 121°C, with a holding time of 20min. The results showed that the come-up time (CUT) for the sterilizer to reach the sterilization temperature at 110°C was 32min, while at 121°C temperature, it required 36min. The lethal rate at a temperature of 121°C reached 1 after 44min of heating, while at a temperature of 110°C, it reached 1 after 40min. The results showed that the fitting between the experimental sterilizer temperature, the cold spot in the cans, and the predicted lethal rate by ANN was good. Moreover, the MSE value was very low, ranging between 6.4152×10^-7 and 0.0034599. The general model Sin 4-8 equations were used to describe the sterilizer temperature, the cold spot in the cans, and the lethal rate, respectively. The results showed that the thermal penetration coefficient (f_h) was 15.01 and 28.57 at temperatures of 110 and 121°C, respectively. The TBA value in sterilized canned fish at temperatures of 110 and 121ºC was 0.11 and 0.16, respectively. The predicted shelf life was 15.83 and 18.64 months at temperatures of 110 and 121°C, respectively.