交互多模型Interacting Multiple Model，IMM）方法应用于高频地波雷达（High Frequency Surface Wave Radar，HFSWR）机动目标跟踪时，其固定的模型集合与转移概率矩阵会导致模型竞争与模型概率切换滞后，降低了目标状态估计的准确性，容易引起航迹断裂、丢失等问题。为此，对基于航道辅助的交互多模型地波雷达机动目标跟踪方法进行了研究。首先，利用有向图模型估计海上航道并提取航道信息，判定目标所处的航道位置并据此自适应切换目标运动模型集合；然后依据确定的模型集合进行多模型滤波，利用滤波新息构造运动模型的极大似然函数并计算每个模型的似然值，通过计算运动模型之间似然值比值计算概率修正因子，采用概率修正因子与原模型转移概率矩阵的乘积更新模型转移概率矩阵。利用仿真与实测数据开展了海上机动目标跟踪实验，结果表明，与传统IMM算法相比，本文方法可显著提升模型概率的切换速度，跟踪得到的航迹维持时间提高了54.2%，算法运行时间减少了30%，运动模型匹配平均概率提高了14.1%，实现了对海上目标的长时跟踪。
When the interactive multiple model method is applied to high frequency surface wave radar maneuvering target tracking, its fixed model set and transition probability matrix will lead to model competition and model probability switching lag, reduce the accuracy of target state estimation, and easily cause track breakage, loss and other problems. In this paper, the method of maneuvering target tracking based on channel aided interactive multi model surface wave radar is studied. Firstly, the directed graph model is used to estimate the sea channel and extract channel information. According to the proposed criteria, the channel position of the target is determined and the target motion model set is adaptively switched; Then, multi model filtering is carried out according to the determined model set. The maximum likelihood function of the motion model is constructed by using the filtering innovation and the likelihood value of each model is calculated. The probability correction factor is obtained by calculating the likelihood value ratio between the motion models and the model transition probability is updated by using the product of the probability correction factor and the original model transition probability. The sea maneuvering target tracking experiment is carried out by using simulation and measured data. The results show that compared with the traditional IMM algorithm, the method in this paper can significantly improve the switching speed of model probability, the track maintenance time obtained by tracking is increased by 54.2%, the algorithm running time is reduced by 30%, and the average matching probability of the moving model is increased by 14.1%, thus realizing the long-term tracking of the sea target.
2023,45(24): 152-159 收稿日期：2022-12-02
DOI：10.3404/j.issn.1672-7649.2023.24.028
分类号：U666.12
基金项目：国家自然科学基金资助项目（62071493，61831010）
作者简介：孙伟峰(1982-),男,博士,副教授,研究方向为高频地波雷达海上目标探测与跟踪
参考文献：
[1] 纪永刚, 张杰, 王祎鸣, 等. 紧凑型高频地波雷达目标探测研究进展[J]. 中国海洋大学学报(自然科学版), 2017, 47(2): 1-7
JI Y G, ZHANG J, WANG Y M, et al. An overview of target monitoring with Compact HFSWR[J]. Periodical of Ocean University of China, 2017, 47(2): 1-7
[2] YE L, WANG Y, YANG Q, et al. High frequency surface wave radar detector base on log-determinant divergence and symmetrized log-determinant divergence[J]. Journal of Electronics & Information Technology, 2019, 41(8): 1931-1938
[3] VIVONE G, BRACA P, HORSTMANN J. Knowledge-based multitarget ship tracking for HF surface wave radar systems[J]. IEEE Transactions on Geoscience and Remote Sensing, 2015, 53(7): 3931-3949
[4] BENAVOLI A, L CHISCI, A FARINA S, et al. Knowledge-based system for multi-target tracking in a littoral environment[J]. IEEE Transactions on Aerospace and Electronic Systems, 2006, 42(3): 1100-1119
[5] SUN M, MA Z, Li Y. Maneuvering target tracking using IMM kalman filter aided by elman neural network[C]//2015 7th International Conference on Intelligent Human-Machine Systems and Cybernetics, 2015, 44-148.
[6] GUO J P. An improved IMM algorithm based on maneuvering-adaptive model set[C]//2016 CIE International Conference on Radar (RADAR), 2016, 1-5.
[7] HAN B, HUANG H, LEI L, et al. An improved IMM algorithm based on STSRCKF for maneuvering target tracking[J]. IEEE Access, 2019, 7: 57795-57804
[8] CLOUTIER J R, LIN C-F, YANG C. Enhanced variable dimension filter for maneuvering target tracking[J]. Aerospace & Electronic Systems IEEE Transactions on Aerospace and Electronic Systems, 1993, 29(3): 786-797
[9] XIE G, SUN L, WEN T, et al. Adaptive Transition Probability Matrix-Based Parallel IMM Algorithm[J]. IEEE Transactions on Systems, Man, and Cybernetics:Systems, 2021, 51(5): 2980-2989
[10] FORTI N, MILLEFIORIL BRACA. M, P. Unsupervised extraction of maritime patterns of life from Automatic Identification System data[C]//OCEANS 2019 - Marseille, 2019, 1-5.
[11] MA J, GUO X. Combination of IMM algorithm and ASTRWCKF for maneuvering target tracking[J]. IEEE Access, 2020, 8: 143095-143103
[12] HWANG I, SEAH C E. A Study on stability of the interacting multiple model algorithm[J]. IEEE Transactions on Automatic Control, 2017, 62(2): 901-906
[13] 夏楠, 王珏, 李博. 基于粒子滤波和交互多模型的移动定位方法[J]. 电子学报, 2019, 47(1): 197: 203.
XIA N, WANG J, LI B. A Mobile localization method based on particle filter and interacting multiple models[J]. Acta Electponica Sinica, 2019, 47(1): 197: 203.
[14] 戴定成, 姚敏立, 蔡宗平, 等. 改进的马尔可夫参数自适应IMM算法[J]. 电子学报, 2017, 45(5): 1198-1205
DAI D C, YAO M L, CAI Z P, et al. Improved adaptive markov IMM algorithm[J]. Acta Electponica Sinica, 2017, 45(5): 1198-1205
[15] 王平波, 刘杨. 基于改进自适应IMM-UKF算法的水下目标跟踪[J]. 电子与信息学报, 2022, 44(6): 1999-2005.
WANG P B, LIU Y Underwater Target Tracking algorithm based on improved adaptive IMM-UKF[J]. journal of electronics & information technology, 2022, 44(6): 1999-2005.
