基于时空数据融合理论的冷链温度监控系统

A Data Fusion Methodfor Cold Chain Temperature Monitoring System

针对易腐农产品冷链物流过程中温度传感器所采集的大量数据,提出时间-空间数据融合理论,并将之应用于易腐农产品冷链物流温度监控系统中,使采集数据能准确地反映冷藏车厢运行状况。首先,对单传感器采用分批估计方法,按照时间先后对观测值进行数据融合,得到每个传感器的局部温度估计值;然后,对单个传感器局部估计值采用自适应调节各传感器权重的方法,对位于空间不同位置的传感器进行数据融合,从而得到冷链车厢监测温度的最终融合值,以期获得比有限个传感器的算术平均值更准确和更可靠的测量结果,以提高信息收集的效率,实现网络节能。数据分析表明,处理后的数据更接近测量真实值。

This paper presents a new data fusion algorithm based on time-space fusion theory for perishable food cold chain monitoring system. Firstly, the data of single sensor is fused according to the time order by means of batch estima- tion method in order to get the partial temperature value. Then, the variance of partial temperature value and the final fu- sion valve is used to adaptively adjust the weight of each sensor, and the ultimate fusion value is obtained. The algorithm can make the performance calculation result reflect the refrigerator van＇ s real status well and truly, which is simple and reliable. Compared with arithmetic means of finite sensors, it can obtain more exacter measuring result. The algorithm can also reflect redundancy or complementary information of multi-sensor in space or time.