Kinetic and Artificial Neural Network Modeling of Dried Black Beauty Eggplant (solanum Melongena L.) Slices During Rehydration

This paper targeted to investigate rehydration characteristics of dehydrated black beauty eggplant (which was dried under sun, by hot air, microwave, and infrared) at 25, 50, and 75 degrees C, and to analyze rehydration ratio (RR) by artificial neural network (ANN) approach. RRs of both microwave-processed specimens (between 5.66 and 7.91) and sun-dried ones (between 5.38 and 6.37) were always greater than the others, at all rehydration temperatures. Rehydration temperature had an enhancing effect on RRs, and the image that closely resembled the fresh sample was captured in the microwave-dried rehydrated samples. Nearly 180-240 min was adequate for rehydration of all slices. Zeroth- and first-order kinetic models, Peleg, Peppas, and two-term exponential decay models, as well as a new rational model were tested to describe rehydration kinetics. Two-term equation was almost superior with high R-2 (between 0.9734 and 0.9994) and the lowest root mean square error (RMSE) and chi(2). Novel recommended mathematical expression was also successful (R-2: 0.9679-0.9989, chi(2): 0.0051-0.0975, RMSE: 0.0191-0.0866). Regarding relationship between actual and predicted RRs and performance indices of ANN equation, overall R and R-2 were recorded as follows: 0.9975-0.9950 (sun-dried) > 0.99642-0.9929 (hot air-dried) > 0.9955-0.9911 (infrared-dried) > 0.9907-0.9815 (microwave-dried), respectively. The proposed ANN model and novel mathematical formula not only offer a considerable potential in predicting rehydration patterns and developing rehydration protocols in food production sector, but also contribute to save energy by completely understanding the process and optimizing conditions.