一种改进的双基地声纳马尔可夫定位算法

A Modified Markov Localization Algorithm of Bistatic Sonar

借鉴岭估计的思想,通过对加权矩阵特征值基准修正,提出了一种改进的马尔可夫定位算法,实现了测量误差协方差阵逆的全域性存在,解决了马尔可夫估计算法在双基地声纳定位应用中的奇点问题.利用该算法对某双基地声纳系统进行了针对定位精度的几何分布的实例仿真,定量分析利比较了3种不同组合方式下算法的定位精度.仿真实例表明,与最佳线性数据融合定位算法相比,运用改进的马尔可夫定位算法对水下目标进行定位可以获得更高的定位精度,同时在基线附近和发射站接收站两侧区域内定位性能亦有较大的提高.算法拓展了冗余数据组合估计在声纳系统中的应用范围.

A modified Markov localization algorithm is given based on the eigenvalue reference correction of the weighting matrix by accepting the idea of the ridge estimation method. The global existence of inverse of measurement errors＇ covari- ance matrix is realized and the singular point problem of Markov estimation algorithm in the application of bistatic sonar are settled by the algorithm. The simulation for some bistatic sonar system is done focusing on the location accuracy geometric distribution. The location accuracy of three combination patterns is analyzed quantitatively and compared. The simulation results show that compared with the localization algorithm based on optimum linear data fusion, the accuracy of locating the underwater objects and the ability of locating the region near the baseline and the two base stations are both improved greatly by the modified Markov localization algorithm. It extends the application range of redundant data combinational estimation in sonar system.