论文部分内容阅读
在分析粗差对Kalman滤波器性能影响的基础上,通过将滤波新息的加权方式改进为深度加权平均,提出了一种基于Kalman框架的新型的稳健滤波算法.该算法仅需引入一个样本深度及权函数的计算步骤,无需针对测元的粗差检择,直接调节各测元对滤波状态的贡献.深度加权滤波扩展了传统Kalman滤波的最小均方误差优化准则,充分利用了不同测元间的相关性和测元与状态的相关性,可以有效降低含粗差数据对滤波结果的影响程度.在稳健性分析的基础上,数值算例验证了算法的可行性和有效性.
Based on the analysis of the effect of gross error on the performance of Kalman filter, a new robust filtering algorithm based on Kalman framework is proposed by improving the weighted method of filtering interest to depth-weighted averaging, which requires only one sample depth And the weighting function step, without any need for the gross error of the measurement unit to directly adjust the contribution of each measurement element to the filtering state.Deep weighting filtering expands the minimum mean square error optimization criterion of traditional Kalman filtering, makes full use of different measurement elements The correlation between them and the correlation between the element and the state can effectively reduce the influence of the data containing the gross errors on the filtering results.On the basis of the robustness analysis, numerical examples show the feasibility and effectiveness of the algorithm.