CAJ | 학술논문

针对机动目标跟踪问题,采用扩维的匀速和转弯(ACV-2ACT)模型组成模型集,结合电扫描雷达波束灵活的特点,利用自适应调度算法,设计了一种基于UT变换的自适应扩维IMM算法.通过无味滤波器估计转弯率和状态,使模型集具有自调整能力.与EKF-AIMM算法进行仿真比较,结果表明在相同情况下,本算法有较好的稳定性,能更加有效节约雷达资源.
침대궤동목표근종문제,채용확유적균속화전만(ACV-2ACT)모형조성모형집,결합전소묘뢰체파속령활적특점,이용자괄응조도산법,설계료일충기우UT변환적자괄응확유IMM산법.통과무미려파기고계전만솔화상태,사모형집구유자조정능력.여EKF-AIMM산법진행방진비교,결과표명재상동정황하,본산법유교호적은정성,능경가유효절약뢰체자원.
For maneuvering target tracking, a new augmented interacting multiple model(AIMM) technique based on unscented transformation is developed in this paper. The model set of this multiple model algorithm is established in terms of one augmented constant velocity (ACV) model and two augmented coordinated turn (ACT) models. Combined with an adaptive revisit concept, the method is applied to electronically scanned radars. The tam rates and the state estimation process are realized by unscented Kal-man filter and they ensure significant self-adjnsting capabilities. Compared with EKF-AIMM algorithm, it is shown that the ap-proach possesses better stability, and can economize radar resource more efficiently.