节点文献
多传感器数据融合方法在系统建模中的应用
The Application of Multi-Sensor Data Fusion in System Modeling
【摘要】 阐述了多传感器数据融合建模的几类典型的方法,重点分析了基于传统多元统计理论的分析法、最小二乘法、人工神经网络等方法的基本原理和各自的技术特点。同时,较为详细地介绍了一种新颖、有效的数据融合建模方法——偏最小二乘回归分析方法(PLSR),并在文章结尾给出了一个运用该方法分析刀具磨损量的实例。
【Abstract】 The typical methods of the multi-sensor data fusion modeling are introduced in this article. It describes the principle and characteristic of the multiple statistic theory, least-square and artificial nerve net etc. in details. Meanwhile, the paper provides a new and effective method of multi-sensor fusion modeling that is called Partial Least Square Regression (PLSR) and its application in analysis of cutter wear is given in the last of the article.
【关键词】 数据融合;
偏最小二乘(PLS);
主成分分析;
神经网络;
【Key words】 data fusion; PLSR; principal component analysis; ANN;
【Key words】 data fusion; PLSR; principal component analysis; ANN;
【基金】 国家自然科学基金资助项目(编号:59975008)
- 【文献出处】 系统仿真学报 ,Acta Simulata Systematica Sinica , 编辑部邮箱 ,2001年S1期
- 【分类号】TP14
- 【被引频次】13
- 【下载频次】314