基于相似度数据融合的车辆航向角研究

Vehicle Heading Angle Research Based on Similarity Data Fusion

为提高车辆行驶过程中的轨迹和姿态控制精度,需将同质车载传感器的测量值进行数据融合,以此获取高精度的关键控制变量。将司南高精度定位、VBox和车辆CAN总线所测得的车辆航向角数据,使用基于相似度的数据融合算法,对测量结果进行数据融合,并分别计算各个数据融合源、数据融合最终值与选定的参考目标值之间的误差。结果表明,数据融合最终值的误差限为0.23°,分别小于数据融合源的0.8°、2.1°和2.7°,精度均高于其它数据融合源。

In order to improve the accuracy of the vehicle trajectory and attitude control during driving,it is necessary to integrate the measured data from the homogeneous sensors to obtain the high-precision key control variables.The measured values of the heading angle from the CM510-21 T,VBox and the vehicle CAN bus were fused by using the similarity-based data fusion algorithm. The errors in the data fusion sources and between the final value after data fusion and the referential target value were calculated. The results show that the margin of error for the final value is 0.23°,which is lower than 0.8°,2.1°,2.7° originating from the data fusion sources,and the accuracy after data fusion is higher than that from each data fusion source.