CLASSIFICATION OF THE ENERGY AND EXERGY OF MICROWAVE DRYERS IN DRYING KIWI USING ARTIFICIAL NEURAL NETWORKS

This investigation uses the artificial neural network model to classify theenergy and exergy of the kiwi drying process in a microwave dryer. In thisexperiment, classification was carried out separately for various pretreatments and microwave powers using three pretreatments (oven, ohmic,and control treatments) and microwave power values (360, 600, and 900W),and the artificial neural network model. Classification was done using 5different input data groups. The first group included the overall data (energyefficiency, special energy loss, exergy efficiency, and exergy loss), while thesecond to fifth groups included the data on the exergy efficiency, specialenergy loss, energy efficiency and special exergy loss in the ordermentioned, which served as the classification inputs. Considering the results,the best R and Percent Correct values for the oven (Percent Correct=90 –R=0.709) and ohmic (Percent Correct=83.33– R=0.846) pretreatments wereobtained. The values of this parameters were also calculated for the control(Percent Correct=71.43 – R=0.843), the 360W power (PercentCorrect=92.86 – R=0.9975), the 600W power (Percent Correct=100 –R=0.9124), and the 900W power (Percent Correct=100 – R=0.9685). Theoverall data was used in the classification phase. In addition, the maximumcorrectly detected data for the oven, ohmic, and pretreatment was 18 (20items), 15 (18 items), and 5 (7 items), respectively. The maximum correctlydetected data for the 360W power, 600W power, and 900W power levelswas 13 (14 items), 15 (15 items), and 16 (16 items), respectively. In sum,the neural network using the overall data input displayed acceptableefficiency in classifying the energy and exergy data of the kiwi dryingprocess in microwave dryers