基于地磁的车型识别模糊数据融合方法研究

把由各向异性磁阻传感器检测获得的地磁曲线视为车辆磁偶极模型的映像，通过分析车辆地磁曲线的峰值、峰谷比、峰值时间、谷值时间、车身长等特征的选择和提取，建立相应车型特征向量；然后，依据各类车型特征，应用三角形隶属度函数特征描述和模糊数据融合，提出了一种基于地磁传感器实时数据采集的车辆检测与车型识别流程和车型识别分类算法；最后，通过一组实际采集的车辆检测数据的分类结果，比较了4种模糊数据融合算式的差异，证明了车辆识别分类算法的有效性，并讨论了有待进一步研究的问题。

Research of Fuzzy Data Fusion Based on Geomagnetism for Vehicle Type Identification Process

Firstly, the curve of the vehicle detected using AMR sensor was mapped to a magnetic dipole. By analysis of the choosing and eliciting the peak, vale, peak time and peak-vale rate of the curve, appropriate feature vectors were established. And then, a vehicle classification algorithm based on geomagnetic sensor real-time data collection was realized on the basis of fuzzy data fusion, the feature is distilled according to various feature classes of vehicle types and depicted by using of triangular function. Finally, four data fusion functions for each type of vehicle were presented and compared to testify the effectiveness, and the issues in future workings were also discussed.