Characterization of Paddy Rice under Different Drying Conditions and Modeling of Moisture Prediction

WU Lan; LIU Huan; SHANG Qingsong

doi:10.13982/j.mfst.1673-9078.2026.1.1436

Abstract：To enhance the quality of rice during the drying process and accurately predict moisture variations,Japonica rice variety was evaluated using fissure-radio and drying time as metrics.The drying process was optimized using single-factor and orthogonal experiments.The variations in the moisture content and quality of rice under different drying temperatures,wind speeds,and initial moisture levels were examined.A genetic algorithm-long short-term memory(GA-LSTM)model integrated with adaptive mutation and elite strategy optimization(AEO)was proposed to predict the moisture content during rice drying.Drying temperature and wind speed significantly affected the fissuring rate and drying time of rice(P<0.01),with their influence ranked as follows:drying temperature>wind speed>initial moisture content.As temperature and wind speed increased,drying rates accelerated(P<0.01),resulting in a significant increase in the fissuring rate.The time-series prediction performances of the BP,LSTM,GA-LSTM,and AEO-GA-LSTM models were constructed and compared under varying drying conditions.The improved AEO-GA-LSTM model achieved a coefficient of determination(R²)of 0.997 0 and a root mean square error(RMSE)of 0.08,outperforming the BP,LSTM,and GA-LSTM models,which had RMSEs of 0.22,0.19,and 0.14,respectively,thus demonstrating superior model fit and reliability.In addition,the improved model exhibited better timeliness by enhancing the computational efficiency by 48.82%and 13.33%compared with the BP and GA-LSTM models,respectively.Therefore,the moisture prediction model established in this study provides a novel approach and methodological reference for moisture prediction under hot-air drying conditions,contributing to the improvement of automation and quality control in rice drying processes.