There are two types track forecast technologies at present. One is model-free type technologies which have little theoretical support, limited capacity and narrow scope of application. The other is model type technologies which have too many prior hypothesis, strict prerequisite conditions and poor universality. Against above problems and to solve the track forecast problem effectively, an uncertain track adaptive forecast model and corresponding exemplary implementation are proposed based on the structures of recurrent neural network and the multi-layer neural network. The proposed model has rigorous theoretical support, less a priori hypotheses, wide application range and strong versatility, which inherits benefits and overcomes weaknesses of existing track forecast methods. Simulation and experimental results show that the proposed model can extract and recognize the patterns in the data set, and make correct and effective prediction according to the recognized pattern, significantly solving the track forecast problems in real environments.
[1] 郭运韬, 朱衍波, 黄智刚. 民用飞机航迹预测关键技术研究[J]. 中国民航大学学报, 2007, 25(1):20-24. GUO Y T, ZHU Y B, HUANG Z G. Study on key trajectory prediction techniques of civil aviation aircraft[J]. Joural of Civil Aviation University of China, 2007, 25(1):20-24(in Chinese).
[2] WILSON R C, WHITLEY T D, ESTKOWSKI R. Trajectory prediction:USA, US20060224318[P]. 2006.
[3] ALAHI A, GOEL K, RAMANATHAN V, et al. Social LSTM:Human trajectory prediction in crowded spaces[C]//Computer Vision and Pattern Recognition. Piscataway, NJ:IEEE Press, 2016:961-971.
[4] 邸忆, 顾晓辉, 龙飞. 基于灰色残差修正理论的目标航迹预测方法[J]. 兵工学报, 2017, 38(3):454-459. DI Y, GU X H, LONG F. Target track prediction method based on grey residual modification theory[J]. Acta Armamentarii, 2017, 38(3):454-459(in Chinese).
[5] 谭伟, 陆百川, 黄美灵. 神经网络结合遗传算法用于航迹预测[J]. 重庆交通大学学报(自然科学版), 2010, 29(1):147-150. TAN W, LU B C, HUANG M L. Track prediction based on neural networks and genetic algorithm[J]. Journal of Chongqing Jiao Tong University(Natural Science), 2010, 29(1):147-150(in Chinese).
[6] 钱夔, 周颖, 杨柳静, 等. 基于BP神经网络的空中目标航迹预测模型[J]. 指挥信息系统与技术, 2017, 8(3):54-58. QIAN K, ZHOU Y, YANG L J, et al. Aircraft target track prediction model based on BP neural network[J]. Command Information System and Technology, 2017, 8(3):54-58(in Chinese).
[7] FANG W, ZHENG L. Rapid and robust initialization for monocular visual inertial navigation within multi-state Kalman filter[J]. Chinese Journal of Aeronautics, 2018, 31(1):148-160.
[8] 赵洲, 黄攀峰, 陈路. 一种融合卡尔曼滤波的改进时空上下文跟踪算法[J]. 航空学报, 2017, 38(2):269-279. ZHAO Z, HAUNG P F, CHEN L. A tracking algorithm of improved spatio-temporal context with Kalmans filter[J]. Acta Aeronautica et Astronautica Sinica, 2017, 38(2):269-279(in Chinese).
[9] LU Z Y, BA B, WANG J H, et al. A direct position determination method with combined TDOA and FDOA based on particle filter[J]. Chinese Journal of Aeronautics, 2018, 31(1):161-168.
[10] HAN H Z, WANG J, DU M Y. GPS/BDS/INS tightly coupled integration accuracy improvement using an improved adaptive interacting multiple model with classified measurement update[J]. Chinese Journal of Aeronautics, 2018, 31(3):556-566.
[11] 翟岱亮, 雷虎民, 李炯, 等. 基于自适应IMM的高超声速飞行器轨迹预测[J]. 航空学报, 2016, 37(11):3466-3475. ZHAI D L, LEI H M, LI J. Trajectory prediction of hypersonic vehicle based on adaptive IMM[J]. Acta Aeronautica et Astronautica Sinica, 2016, 37(11):3466-3475(in Chinese).
[12] 张翔宇, 王国宏, 李俊杰, 等. 临近空间高超声速滑跃式轨迹目标跟踪技术[J]. 航空学报, 2015, 36(6):1983-1994. ZHANG X Y, WANG G H, LI J J, et al. Tracking of hypersonic sliding target in near-space[J]. Acta Aeronautica et Astronautica Sinica, 2015, 36(6):1983-1994(in Chinese).
[13] 吴楠, 陈磊. 高超声速滑翔再入飞行器弹道估计的自适应卡尔曼滤波[J]. 航空学报, 2013, 34(8):1960-1971. WU N, CHEN L. Adaptive Kalman filtering for trajectory estimation of hypersonic glide reentry vehicles[J]. Acta Aeronautica et Astronautica Sinica, 2013, 34(8):1960-1971(in Chinese).
[14] CHO K, VAN MERRIENBOER B, GULCEHRE C, et al. Learning phrase representations using RNN encoder-decoder for statistical machine translation[J]. Computer Science, 2014(6):1-4.
[15] MAO J, XU W, YANG Y, et al. Deep captioning with multimodal recurrent neural networks(m-RNN)[J]. Eprint Arxiv, 2014:1-15.
[16] CHEN S H, HWANG S H, WANG Y R. An RNN-based prosodic information synthesizer for Mandarin text-to-speech[J]. IEEE Transactions on Speech & Audio Processing, 1998, 6(3):226-239.
[17] LECUN Y, BENGIO Y, HINTON G. Deep learning[J]. Nature, 2015, 521(7553):436.
[18] SCHMIDHUBER J R. Deep learning in neural networks[M]. Amsterdam:Elsevier Science Ltd., 2015.
[19] SCHMIDHUBER J R. Deep learning in neural networks:An overview[J]. Neural Netw, 2015, 61:85-117.
[20] TORRES J F, TRONCOSO A, KOPRINSKA I, et al. Deep learning for big data time series forecasting applied to solar power[C]//International Joint Conference SOCO'18-CISIS'18-ICEUTE'18. Berlin:Springer, 2019.
[21] CHEN S, WEN J, ZHANG R. GRU-RNN based question answering over knowledge base[C]//Knowledge Graph and Semantic Computing:Semantic, Knowledge, and Linked Big Data. Berlin:Springer, 2016:80-91.
CJA