Drying kinetics of tea using artificial neural network model

Drying of the tea leaves for longer period of storage is always required. Tea green leaves in air temperature range from 35 to 55 degrees Celsius and input air velocities 0.5 and 0.7 meters per second and time range 0 to 140 minutes were dried in an experimental dryer process. For this purpose, four samples were considered for each temperature. In each experiment, total weight changes of samples were recorded continuously. Tea drying process by neural network method with four input vector (time, temperature, velocity and moisture) and one output vector (moisture ratio) has been modeled. The obtained Results of neural network were compared with the experimental data.The best result obtained to feed forward back propagation neural network with Leven berg- Marquardt algorithm and TANSIG threshold function which indicates high accuracy of the neural network. The mean relative percent error and correlation coefficient and root mean square error are1.3, 0.9998, 0.00008, respectively.