BP neural network prediction of the effects of drying rate of fruits and vegetables pretreated by high-pulsed electric field

The pretreatment of fruits and vegetables by high-pulsed electric field (HPEF) can increase the drying rate, improve the quality of dried product and reduce energy consumption. Through the HPEF pretreatment drying test of radishes and apples, the effects of pretreatment factors, such as pulse intensity, action time and pulse number on drying rate, were investigated. This paper analyzed the relationship between HPEF parameters and drying rate at different periods based on BP neural network of L-M training method, thus established the causation between them. The results show that the effect of pulse intensity on drying rate is more significant than the effects of action time and pulse number. The simulation model of BP neural network for pretreatment drying of radishes and apples was built and compared with the measured values. The drying rate predicted by the BP neural network is close to the measured values, so BP neural network can be used to estimate the result.