Modeling of moisture content variation under variable microwave power in microwave-hot air dryer is challenging. In this study, one static and one dynamic of ANN were investigated for modeling of whole mushroom drying process. The experiments were done in three levels of hot air temperature (23, 50 and 70 °C), three levels of microwave power density (MPD) (1.5, 2 and 2.5 W/g) and two statuses of microwave power during drying process (constant and variable). The results demonstrated that the two distinct falling rate periods were observed at low MPD with any air temperatures and power density status. The lowest specific energy consumption was found in 23°C and 2.5 fixed MPD and lowest color deterioration found in 70 °C and 1.5 fixed MPD. The results also showed the 3-6-7-1 structure of dynamic model with 0.2179 RMSE and 0.9914 R values had better results than other structures and would be very suitable in a predictive control system for drying control.
