CAJ | 학술논문

针对舰机传递对准中的大姿态失准角问题，研究了稀疏高斯厄米特积分滤波在舰机传递对准中的应用。为了解决稀疏高斯滤波中积分点利用效率不高的问题，提出了一种改进的自适应各向异性稀疏高斯积分滤波算法，改进算法根据状态参数可观测度大小将积分变量分类并根据分类结果构造各向异性权重向量来控制各维通道积分的精度等级，进而对各变量通道的积分点数目进行合理分配。模拟轨迹下仿真结果表明，采用稀疏高斯滤波与采用改进算法的舰机传递对准精度相当，较无味卡尔曼滤波有所提高，对准精度由3?提高到2?；改进算法的积分点数目减少了30%~60%，提高了计算效率。
침대함궤전체대준중적대자태실준각문제，연구료희소고사액미특적분려파재함궤전체대준중적응용。위료해결희소고사려파중적분점이용효솔불고적문제，제출료일충개진적자괄응각향이성희소고사적분려파산법，개진산법근거상태삼수가관측도대소장적분변량분류병근거분류결과구조각향이성권중향량래공제각유통도적분적정도등급，진이대각변량통도적적분점수목진행합리분배。모의궤적하방진결과표명，채용희소고사려파여채용개진산법적함궤전체대준정도상당，교무미잡이만려파유소제고，대준정도유3?제고도2?；개진산법적적분점수목감소료30%~60%，제고료계산효솔。
In view of the problems of large misalignment angles in carrier-aircraft INS’s transfer alignment, a sparse Gauss-Hermite quadrature filter(SGHQF) is applied to transfer alignment. To solve the problems of low utilization efficiency of points in SGHQF, a new adaptive anisotropic sparse Gauss-Hermite quadrature filter(AASGHQF) is proposed. The state parameters’ observable degree is quantized to classify the state parameters, and a mechanism for controlling the accuracy level of each dimension is presented by structuring adaptive anisotropic importance vector, in which the value of every element is determined according to grouped results. This mechanism can distribute the quadrature points nonuniformly and reasonably. Simulation results show that the AASGHQF algorithm has comparative performance with SGHQF and has a better performance compared with Unscented Kalman filter(UKF). The alignment precision of 3 arcmin could be improved to 2 arcmin by this algorithm. In addition, the number of AASGHQ points is decreased by 30%~60%and the computation efficiency is increased.