基于D-S证据理论的柑橘环境监测数据融合研究

Research on Citrus Environmental Monitoring Data Fusion Based on D-S Evidence Theory

针对果园监测数据量大、采集数据存在不确定性的问题,结合柑橘生长特点,提出一种应用于柑橘环境监测的基于D-S证据理论的多传感器数据融合方法,可有效划分同类、异类数据,提取同类数据特征值,计算数据间距,剔除异常值,得到初步融合数据,进而计算异类数据与各等级特征间的距离,构建概率分配函数,从而获取精确的柑橘环境监测数据。结果表明,该方法可以得到较为准确的融合结果,具有一定的可拓展性。

Aiming at the problem of large amount of orchard monitoring data and uncertainty of collected data, combined with the growth characteristics of citrus, a multi-sensor data fusion method based on D-S evidence theory is proposed for citrus environmental monitoring, which can effectively divide the same kind of data and heterogeneous data, extract the characteristic values of the same kind of data, calculate the data spacing, eliminate the abnormal values, obtain the preliminary fusion data, and then calculate the distance between heterogeneous data and the characteristics of each grade. The probability distribution function is constructed to obtain accurate citrus environmental monitoring data. The results show that this method can get more accurate fusion results, and has certain scalability.