多传感器数据采集系统中的数据融合研究

针对多传感器数据采集系统的准确性与可靠性的要求,提出了单传感器分批估计融合与模糊理论中的相关性函数与加权自适应算法相结合的数据融合方法。首先,对每一只传感器进行分批估计融合,得到了较为可靠的局部决策值。然后,利用模糊理论中的相关性函数,计算出每只传感器与其它传感器的相关程度,并剔除掉那些相关程度较低的传感器的局部决策值。最后,在权的最优分配原则下,利用加权自适应算法对剩下的局部决策值进行数据融合。该方法能在数据先验知识未知的情况下,客观地反映各个传感器的可靠程度。数据分析结果表明:与同类融合方法相比,该方法具有更高的精度。

Study on data fusion of multi-sensor data acquisition system

Aiming at requirements of accuracy and reliability of multi-sensor data acquisition system,the data fusion method that combines batch estimation fusion of single sensor with the correlation function of fuzzy theory and the adaptive weighted fusion algorithm is proposed.Firstly,batch estimation fusion is applied to every single senor to obtain more reliable value of local decision-making.Secondly,correlation function of fuzzy theory is applied to calculate the degree of correlation of each sensor with others,then weed out the local decision-making value of low degree of correlation.At last,the adaptive weighted fusion algorithm is applied to the remaining value of local decision-making according to the principle of optimal weight distribution for data fusion.This method objectively reflects the reliability of every sensor without any priori knowledge.Data analysis result shows that this method has higher precision,compared with fusion method of similar.