目的 更好地满足车辆自动驾驶时航向角测量的精度要求。方法 提出卡尔曼滤波算法，把实时动态-全球导航卫星系统（RTK-GNSS）测量出来的经纬度和高程经过高斯投影转换为高斯平面坐标，和微电子机械系统（MEMS）陀螺仪测得的累积航向角进行融合处理，最终得到车辆更为精准的航向角。结果 融合后的航向角度曲线既保持了GNSS航向的整体变化趋势，也保持了陀螺仪航向的细部变化趋势，且较GNSS和陀螺仪所得曲线更为平滑，可以跟踪车辆180°调头的转弯动作。结论 卡尔曼滤波算法可以实时在线且精准地测得车辆航向角数据，精度较GNSS测量结果提高80%以上。
Objective To better meet the accuracy requirement of vehicle heading angle measure.Method Kalman filter algorithm was proposed. The latitude, longitude and elevation measured by RTK-GNSS receiver were converted to plane coordinates using Gaussian projection. The Gaussian plane coordinates and the accumulated heading angle measured by gyroscope were integrated by Kalman filter, and finally more accurate heading angle was obtained.Result The integrated curve kept the entire variation trend of the heading angle measured by GNSS receiver and the partial variation trend measured by MEMS gyroscope. The curve was smoother than that based on GNSS and gyroscope, and could follow action of vehicle 180° turning.Conclusion The Kalman filter algorithm can measure the vehicle heading angle data in real time and the precision was improved 80% more than the result measured by GNSS.