CAJ | 학술논문

针对强跟踪滤波器( STF)的理论局限以及基于UT变换的强跟踪滤波器( UTSTF)处理高维非线性系统时滤波精确度下降甚至发散等问题，提出一种基于容积卡尔曼滤波( CKF)算法的强跟踪滤波器( CKFSTF)。 CKFSTF兼具了STF和CKF的优点：鲁棒性强，滤波精度高，数值稳定性好，计算速度快，容易实现且应用范围广。此外，对于目标跟踪系统过程噪声统计特性未知的情况，在CKFSTF的基础上应用Sage-Husa噪声估值器对噪声统计特性进行在线估计，形成自适应CKFSTF。仿真结果验证了新算法的有效性。
침대강근종려파기( STF)적이론국한이급기우UT변환적강근종려파기( UTSTF)처리고유비선성계통시려파정학도하강심지발산등문제，제출일충기우용적잡이만려파( CKF)산법적강근종려파기( CKFSTF)。 CKFSTF겸구료STF화CKF적우점：로봉성강，려파정도고，수치은정성호，계산속도쾌，용역실현차응용범위엄。차외，대우목표근종계통과정조성통계특성미지적정황，재CKFSTF적기출상응용Sage-Husa조성고치기대조성통계특성진행재선고계，형성자괄응CKFSTF。방진결과험증료신산법적유효성。
For the problem that Strong tracking filter ( STF) has some theoretical limitations and the STF based on unscented transformation ( UTSTF) declines in accuracy and further diverges when solving the nonlinear filtering problem in high dimension, a cubature Kalman filter ( CKF) with strong tracking be-havior ( CKFSTF ) was proposed. CKFSTF combines advantages of STF and CKF: strong robustness, high accuracy, strong numerical stability, fast calculation speed, easy implementation and wide range of applications. Furthermore, adaptive CKFSTF was proposed when the prior noise statistic is unknown and time-varying, which using Sage-Husa noise statistic estimator based on CKFSTF. Validity of the new pro-posed algorithm was verified by the simulation examples.