| 摘要: |
| 为增强多传感器测量数据融合效果，在综合考虑传感器初始精度与实际测量精度的基础上，提出了一种改进的自适应加权融合算法。将证据理论中的修正证据距离引入传感器实际测量数据间距离计算，基于计算得出的测量数据间距离生成各数据融合时的测量权重值。当传感器精度已知或者能够计算得出时，将基于传感器精度生成的固定权重与测量权重相综合，生成最终权重；当传感器精度未知且无法计算得出时，将测量权重作为最终权重。基于多种典型算例对所提融合算法进行验证，结果表明所提算法融合效果较好，具有一定的理论意义和较好的工程实用价值。 |
| 关键词: 数据融合 自适应加权 修正证据距离 测量权重 固定权重 |
| DOI： |
|
| 基金项目:国家自然科学基金资助项目（51475472，61403396） |
|
| An adaptive weighted data fusion algorithm considering sensor accuracy |
| XING Xiaochen,CAI Yuanwen,REN Jiangtao,ZHAO Zhengyu |
| () |
| Abstract: |
| In order to enhance the effect of multi-sensor data fusion,an improved adaptive weighted fusion algorithm is presented,on the basis of comprehensive consideration of the initial accuracy and the measurement accuracy of the sensor. The modified evidence distance of evidence theory is used to calculate the distance of measured data of sensors,and the measurement weight of sensor data can be obtained based on the data distance. The constant weight can be calculated by the initial accuracy of sensor. When the initial accuracy of sensor is known or can be calculated,constant weight of sensor can be obtained. Then the constant weight and the measurement weight are combined to generate the final weight. If the initial accuracy of sensor can not be got,the measurement weight is used as the final weight. Many kinds of typical examples are used to validate the proposed fusion algorithm,and the result shows that the fusion result is satisfying,and the algorithm has theoretical and practical significance. |
| Key words: data fusion adaptive weighting modified evidence distance measurement weight constant weight |
