摘要
在独特型网络的基础上,设计了一种应用于高速公路上,装载CCD和雷达的车辆多传感器信息融合模型。实时更新的抗体库能提高识别速度,而抗体库多样性则通过抗体间正交亲和度阈值来保证。独特型网络中抗体浓度随着抗原入侵产生变化,最终达到另一平衡状态,并得到融合结果及其可信度。抗体浓度的变化由克隆选择算法实现。
For improving veltiele detecting veracity on free-way, this paper developed a multi-sensor data fusion model based on the Idiotypic Network Theory （INT）. Real-time update antibody repertoire can speed the identification, and its diversity was maintained by fitness calculation between antibodies. The concentration of antibodies in idiotypic net would change continuously as antigen invading till the net reach another balance state. The balance states indicated the fusion results and their reliabilities.
出处
《机械工程师》
2006年第3期74-76,共3页
Mechanical Engineer
关键词
多传感器信息融合
独特型网络
人工免疫系统
汽车
multi-sensor data fusion
idiotypic network theory （int）
artificial immune system
vehicle
