基于抗差Kalman滤波算法的单频BDS周跳探测

Single Frequency Beidou Cycle Slip Detection Based on Robust Kalman Filter

提出了一种基于抗差Kalman滤波算法的单频BDS周跳探测方法。该方法以Kalman滤波算法为基础,构建了历元间差分观测值,使周跳以粗差的形式表现出来,并采用IGGⅢ等价权函数调整滤波增益矩阵实现抗差处理,从而达到准确探测周跳的目的。采用单频BDS实测数据对本方法进行验证,并以现有Kalman滤波方法作为对比。结果表明,在无周跳情况下,Kalman滤波法预报残差序列中误差为0.140,抗差Kalman滤波法预报残差序列中误差为0.121,本方法对观测噪声的灵敏性更强;在加入多个具有不同特征的模拟周跳后,Kalman滤波法存在滤波多次初始化问题而不能探测出少量历元的连续周跳,本方法则实现了所有周跳的探测,准确率为100%,同时保证了滤波过程只需初始化1次。

A single-frequency Beidou cycle slip detection method based on the robust Kalman filter algorithm is proposed.Based on the Kalman filter algorithm,the method constructs the difference observation between adjacent epochs,so that the cycle slip is expressed in the form of gross error,and the IGGIII equivalent weight function is used to adjust the filter gain matrix to complete the robust processing,so as to achieve the purpose of accurately detecting the cycle slip.Single frequency BDS measured data is used to verify the method in this paper,and the existing Kalman filter method is used as a comparison.The results show that in the case of no cycle slip,the RMS of the residual sequence predicted by the Kalman filter method is 0.140,and the RMS of the residual sequence predicted by the robust Kalman filter method is 0.121,so the proposed method is more sensitive to the observation noise.After adding multiple simulated cycle slips with different characteristics,the Kalman filter method has the problem of multiple initializations of the filter and cannot detect continuous cycle slips with a small number of epochs.The proposed method realizes the detection of all cycle slips with an accuracy rate of 100%,while ensuring that the filtering process only needs to be initialized once.