The Prediction Research of Population Density Based on Deep Learning in Grain Stored Insects

Abstract

Precision of pests, in stored grain insect population density, has been a hot and difficult research in pest detection and control system. The accuracy of prediction of pest density will directly affect to warehouse grain temperature and the food quality etc. In order to improve the accuracy, the paper which using the depth study method, established an insects density prediction mode with the depth of the belief network as the core. The model is applied to the algorithm of deep learning predictive control. According to the temperature and humidity of the grain obtained from the actual measurement and the initial density of the pest, we predicted the pest density. Simulation results show that the root mean square error is small between the predictive value and actual value, high prediction accuracy. The deep learning algorithm is applied to the population density of pests is effective.