Paddy Moisture Content and Drying Time Prediction Using Neural Networks

The current paddy drying practice in the rice mill is heating and blowing air continuously to reduce moisture content a certain level (i.e., 12%–14%). stopping time of process determined via manual checking (MC) from time. Several problems inconsistent dried quality excess energy consumption) arise usual due human error. Hence, MC time-series data collected an actual can be related environment factors build profile using Long Short-Term Memory (LSTM) neural networks. prediction carried out every 2 hours up 20 beginning process. Based on level, optimum when reaches desired level. When evaluated against validation dataset, our system successfully predicted 10 ahead time, with Root Mean Square Error (RMSE) just 1.5 units.