摘要
工程领域里的一类目标参数估计常常存在多个估计模型。本文提出一种BP神经网络数据融合算法,对两个模型给出的四个估计结果有效融合,力图给出一个可靠性和精度更高的融合结果。将融合结果与原模型估计结果比对分析发现,融合算法可以有效降低失效率,在一定程度上提高精度。
There often exist several estimation models to a kind of target parameters esumauon in engineering, field. In this article, the data fusion method based on BP neural network is proposed to fuse four outputs of two estimation models and try to produce fusion results with higher reliability and higher precision. Comparing fusion results with original estimations, it is found that the data fusion method can effectively reduce probability of invalidation and improve the precision to a certain extent.
出处
《信息与电子工程》
2005年第3期197-200,共4页
information and electronic engineering
基金
中国工程物理研究院科学技术基金(20030432)
关键词
信息处理技术
数据融合
算法
BP神经网络
可靠性
精度
information processing technology
data fusion
algorithm
BP Neural Network
reliability
precision
