Abstract: The performance of existing maneuvering target tracking methods for highly maneuvering targets in cluttered environments is unsatisfactory. This paper proposes a hybrid-driven approach for tracking multiple highly maneuvering targets, leveraging the advantages of both data-driven and model-based algorithms. The time-varying constant velocity model is integrated into the Gaussian process (GP) of online learning to improve the performance of GP prediction. This integration is further combined with a generalized probabilistic data association algorithm to realize multi-target tracking. Through the simulations, it has been demonstrated that the hybrid-driven approach exhibits significant performance improvements in comparison with widely used algorithms such as the interactive multi-model method and the data-driven GP motion tracker.

基于混合驱动高斯过程学习的强机动多目标跟踪方法

[1]Aftab W, Mihaylova L, 2020. On the impact of different kernels and training data on a Gaussian process approach for target tracking. Proc IEEE 23^rd Int Conf on Information Fusion, p.1-6.

[2]Aftab W, Mihaylova L, 2021. A learning Gaussian process approach for maneuvering target tracking and smoothing. IEEE Trans Aerosp Electron Syst, 57(1):278-292.

[3]Da K, Li TC, Zhu YF, et al., 2021. Recent advances in multisensor multitarget tracking using random finite set. Front Inform Technol Electron Eng, 22(1):5-24.

[4]Deng LC, Li D, Li RF, 2020. Improved IMM algorithm based on RNNs. J Phys Conf Ser, 1518:012055.

[5]Guo YF, Fan KS, Peng DL, et al., 2015. A modified variable rate particle filter for maneuvering target tracking. Front Inform Technol Electron Eng, 16(11):985-994.

[6]Guo YF, Tharmarasa R, Rajan S, et al., 2016. Passive tracking in heavy clutter with sensor location uncertainty. IEEE Trans Aerosp Electron Syst, 52(4):1536-1554.

[7]Guo YF, Li Y, Tharmarasa R, et al., 2019. GP-PDA filter for extended target tracking with measurement origin uncertainty. IEEE Trans Aerosp Electron Syst, 55(4):1725-1742.

[8]Guo YF, Li Y, Ren X, et al., 2020a. Multiple maneuvering extended target tracking based on Gaussian process. Acta Autom Sin, 46(11):2392-2403 (in Chinese).

[9]Guo YF, Li Y, Xue AK, et al., 2020b. Simultaneous tracking of a maneuvering ship and its wake using Gaussian processes. Signal Process, 172:107547.

[10]Guo YF, Zhu JJ, Zhou S, et al., 2022. A joint model and data driven track segment association algorithm for manoeuvring target tracking. IET Radar Sonar Nav, 16(10):1670-1680.

[11]Huber MF, 2014. Recursive Gaussian process: on-line regression and learning. Patt Recogn Lett, 45:85-91.

[12]Li TC, Hlawatsch F, 2021. A distributed particle-PHD filter using arithmetic-average fusion of Gaussian mixture parameters. Inform Fusion, 73:111-124.

[13]Li TC, Su JY, Liu W, et al., 2017. Approximate Gaussian conjugacy: parametric recursive filtering under nonlinearity, multimodality, uncertainty, and constraint, and beyond. Front Inform Technol Electron Eng, 18(12):1913-1939.

[14]Li TC, Liu ZG, Pan Q, 2019a. Distributed Bernoulli filtering for target detection and tracking based on arithmetic average fusion. IEEE Signal Procss Lett, 26(12):1812-1816.

[15]Li TC, Chen HM, Sun SD, et al., 2019b. Joint smoothing and tracking based on continuous-time target trajectory function fitting. IEEE Trans Autom Sci Eng, 16(3):1476-1483.

[16]Liu JX, Wang ZL, Xu M, 2020. DeepMTT: a deep learning maneuvering target-tracking algorithm based on bi-directional LSTM network. Inform Fusion, 53:289-304.

[17]Liu XC, Lyu C, George J, et al., 2022. A learning distributed Gaussian process approach for target tracking over sensor networks. Proc 25^th Int Conf on Information Fusion, p.1-8.

[18]Pan Q, Ye XN, Zhang HC, 2005. Generalized probability data association algorithm. Acta Electron Sin, 33(3):467-472.

[19]Rasmussen CE, Williams CKI, 2006. Gaussian Processes for Machine Learning. MIT Press, Cambridge, USA.

[20]Sun MW, Davies ME, Proudler I, et al., 2020. A Gaussian process based method for multiple model tracking. Proc Sensor Signal Processing for Defence Conf, p.1-5.

[21]Tian WM, Fang LL, Li WD, et al., 2022. Deep-learning-based multiple model tracking method for targets with complex maneuvering motion. Remote Sens, 14(14):3276.

[22]Wang LP, Zhan RZ, Huang Y, et al., 2021. Joint tracking and classification of extended targets with complex shapes. Front Inform Technol Electron Eng, 22(6):839-861.

[23]Wu WH, Cai YC, Jin HB, et al., 2021. Derivation of the multi-model generalized labeled multi-Bernoulli filter: a solution to multi-target hybrid systems. Front Inform Technol Electron Eng, 22(1):79-87.

[24]Xiong W, Zhu HF, Cui YQ, 2022. Recurrent adaptive maneuvering target tracking algorithm based on online learning. Acta Aeronaut Astronaut Sin, 43(5):325250 (in Chinese).

[25]Zhang D, Liu MQ, Zhang SL, et al., 2018. Mutual-information based weighted fusion for target tracking in underwater wireless sensor networks. Front Inform Technol Electron Eng, 19(4):544-556.

[26]Zhang XR, He FH, Zheng TY, 2019. An LSTM-based trajectory estimation algorithm for non-cooperative maneuvering flight vehicles. Proc Chinese Control Conf, p.8821-8826.

[27]Zheng Z, Cai SC, 2021. A collaborative target tracking algorithm for multiple UAVs with inferior tracking capabilities. Front Inform Technol Electron Eng, 22(10):1334-1350.

[28]Zhou R, Feng Y, Bin D, et al., 2020. Multi-UAV cooperative target tracking with bounded noise for connectivity preservation. Front Inform Technol Electron Eng, 21(10):1494-1503.

[29]Zhu Y, Liang S, Wu XJ, et al., 2021. A random finite set based joint probabilistic data association filter with non-homogeneous Markov chain. Front Inform Technol Electron Eng, 22(8):1114-1126.