Abstract: Adhesion and agglomeration easily occur during the traditional drying process of sweet potato residue. In order to improve this phenomenon, the process parameters for the fluidized bed drying of sweet potato residue, i.e., material loading per unit area, bed temperature, air flow rate, were optimized using response surface methodology with a Box-Behnken design. The weighted average of dimensionless drying time and granularity was used as response variable. The results showed that the response was influenced in decreasing order by material loading per unit area, air flow rate and bed temperature. The optimal conditions obtained from the regression model were as follows: loading capacity, 5 264 g/m2; air flow rate, 52.73 m3/h; and bed temperature, 51.33 ℃. Under these conditions, the predicted response value was 0.790, and the validation experiments showed a prediction accuracy of up to 93.60%. Compared with traditional drying method, the time required for fluidized bed drying was shortened. Additionally, 82.3% of the dried product could pass through a 20-mesh sieve, bulk density of the sample was 0.446 g/mL, which was significantly increased by 25.8%, and hardness was 485.382 g, which was significantly reduced. Under scanning electron microscopy, it was observed that the fluidized bed dried product exhibited particles with larger internal voids and had a looser structure, leading to reduced comminution cost. This study may provide a theoretical basis for industrial drying, prolongation of the storage time and intensive processing of sweet potato residue.

Key words: sweet potato residue, fluidized bed, drying, weighted average, quality