基于虚拟多传感器信息融合的粮情预警系统

The Early-Warning System of Grain Circumstances Based on Virtual Multi-Sensor Integration

在粮食仓储过程中,通过温湿度传感器、图像传感器采集到的温湿度、图像的实时数据,经过信息处理算法的变换和特征提取,得到具有不同特征信息的虚拟多传感器数据,采用D-S证据理论对数据进行信息融合,对粮食在仓储过程中的霉变、虫害等粮情进行动态预测分析,当发现有不良变化时,该系统能够及时发出预警信息,以保证粮食的储藏安全。

The temperature, humidity, image data had been obtained by sensors in the grain storage period. The virtual multi-sensor information with different characteristics can be obtained after transformation and feature extraction. Then, the information integration is achieved by Dempster-Shafer evidential theory. Mildew and pests in grain can be forecasted in the grain storage period. When adverse factors occur, this system can send out warning information to ensure grain safety.