Application of Data Fusion on System of Measurement and Control for Grain Storage

Abstract

Data fusion was used to predict the safe value of grain environment in different operational conditions. The temperature, humidity, moisture, pest and etc, were used as inputs to the data fusion center and network. Feed-forward artificial neural networks with 4-10-5 arrangements were capable to estimate the safe value of grain environment as the outputs. This method, characterized by sufficiently utilizing the effective detected data, optimizing homogeneous data, and considering the complementation of the different data source makes possible an improvement in the reliability and entirety of detection system.