Development and optimization of an electrohydrodynamic dehydrator using ANN-GA for improved energy performance

The performance control of an electrohydrodynamic (EHD) dehydrator remains poorly understood. This study experimentally evaluated the drying kinetics and specific energy consumption (SEC) of an EHD drying system. Several parameters influencing the performance of the EHD dehydrator were investigated, including applied voltage (V), mesh size or open area ( %), needle tip sharpness angle (ϴ), inter-needle spacing (D × D), and the distance between the needle tip and the mesh (g). A laboratory-scale EHD dehydrator was developed to study the effects of these variables, using ring-shaped pineapple slices as the test material. The moisture ratio function was derived from the moisture transport equation for the ring-shaped pineapple slices and fitted to the experimental data. The diffusion coefficient extracted from the moisture ratio function was used to assess the drying kinetics, while SEC was used to evaluate the energy efficiency of the EHD dehydrator under different parameter settings. To predict and control performance, an artificial neural network (ANN) model was developed, achieving a high R-value of 0.9781. Furthermore, an integrated ANN–genetic algorithm (ANN-GA) approach was established to optimize the drying parameters, aiming to maximize drying kinetics and minimize SEC. These findings suggest that the ANN and integrated ANN-GA models can be effectively applied to design EHD dehydrators for pineapple drying at a manufacturing scale in the future.