Design and experiment of integrated intelligent equipment for hot air blanching and air-flow impact drying

Air-flow impact drying technology offers high efficiency and good product quality for apple crisp pellets processing but suffers from difficulties in controlling drying conditions and endpoints. To address these deficiencies, this study developed a novel air-flow impact drying equipment (AIDE) and constructed a Bi-LSTM neural network through temperature processing data for real-time monitoring of the moisture content of the material during the drying process. Tests on apple pellets verified the AIDE’s stable and reliable performance; built-in sensors enabled real-time, accurate monitoring of material internal temperature and moisture content. In addition, compared with traditional direct heating, the hot air blanching method adopted in the AIDE showed significant advantages. At the optimal temperature of 75℃, drying time was shortened by 9.69%, vitamin C content increased by 7.7% and color intensity improved by 10.53%. The Bi-LSTM model achieved high monitoring accuracy, with an R² of 0.9998, RMSE of 0.3113 and MAE of 0.2548. This technology can be applied to the processing of multiple types of food. Through the optimization and upgrading of equipment and models, it promotes the development of food processing towards intelligence. At the same time, it helps in the development of high-value-added healthy foods, and has significant potential for market application and industrial upgrading.