基于神经网络的符号化飞行动作识别

doi:10.12305/j.issn.1001-506X.2022.03.04

摘要/Abstract

摘要：

飞行动作识别是飞行训练评估和空战智能决策等多项关键技术的基础, 实现飞行动作的快速高效识别具有重大意义。对此, 提出一种基于神经网络符号化模型的方法, 实现对基本飞行动作和复杂飞行动作高效识别。首先, 利用微分分割的思想对飞行参数进行切片处理, 然后通过卷积神经网络(convolutional neural networks, CNN)和长短期记忆(long-short term memory, LSTM)神经网络实现飞行动作的模块化处理, 有效代替了传统方法中对原始数据的逻辑推理。并且该方法可以利用基本飞行动作对飞行过程实现飞行数据分割, 具有良好的扩展性, 能够快速处理批量飞参数据。最后对13种基本飞行动作、两种复杂飞行动作和整段飞行数据进行仿真实验。仿真结果表明, 该方法具有良好的识别性能。

关键词: 动作识别, 微分分割, 卷积神经网络, 长短期记忆网络

Abstract:

Flight action recognition is the basis of many key technologies, such as flight training evaluation and air combat intelligent decision-making. It is of great significance to realize fast and efficient flight action recognition. Thus, a method based on symbolic neural network is proposed to realize the efficient recognition of basic flight actions and complex flight actions. Firstly, the idea of differential segmentation is used to slice the flight parameters, and then the convolution neural network (CNN) and long-short term memory (LSTM) neural network are used to realize the modular processing of flight actions, which effectively replaces the traditional method of logical reasoning for the original data. And the method can use the basic flight action to segment the flight data. It has good scalability and can process a large number of flight data quickly. Finally, the simulation experiments of thirteen basic flight actions, two complex flight actions and the whole flight data are carried out. The simulation results show that the method has good recognition performance.

Key words: action recognition, differential segmentation, convolution neural network (CNN), long-short term memory (LSTM) network

中图分类号:

V249

方伟, 王玉, 闫文君, 林冲. 基于神经网络的符号化飞行动作识别[J]. 系统工程与电子技术, 2022, 44(3): 737-745.

Wei FANG, Yu WANG, Wenjun YAN, Chong LIN. Symbolized flight action recognition based on neural network[J]. Systems Engineering and Electronics, 2022, 44(3): 737-745.

图/表 22

图1

表1

图2

表2

图3

图4

图5

表3

表4

表5

图6

图7

图8

图9

表6

表7

图10

表8

图11

图12

图13

图14

参考文献 35

1	KELLER J . DARPA eyes artificial intelligence (AI) and unmanned aircraft in jet fighter dogfighting[J]. Military & Aerospace Electronics, 2019, 30 (6): 30.
2	黄长强. 未来空战过程智能化关键技术研究[J]. 航空兵器, 2019, 26 (1): 11- 19.
	HUANG C Q . Research on key technologies of intelligent air combat process in the future[J]. Aeronautical Weapon, 2019, 26 (1): 11- 19.
3	曹建彪, 王业光, 王世鹏, 等. 基于飞行任务的飞行品质评估方法[J]. 飞机设计, 2021, 41 (2): 37- 41.
	CAO J B , WANG Y G , WANG S P , et al. Flight quality assessment method based on flight mission[J]. Aircraft Design, 2021, 41 (2): 37- 41.
4	JING L , QI S J , WANG F . Unmanned aerial vehicle flying qua-lities flight test evaluation technique in approaching and landing phase[J]. IOP Conference Series Materials Science and Engineering, 2020, 887 (1): 12- 42.
5	孟光磊, 张慧敏, 朴海音, 等. 自动化飞行训练评估中的战机机动动作识别[J]. 北京航空航天大学学报, 2020, 46 (7): 1268- 1273.
	MENG G G , ZHANG H M , PIAO H Y , et al. Recognition of flighter maneuver in automatic flight training evaluation[J]. Journal of Beijing University of Aeronautics and Astronsutics, 2020, 46 (7): 1268- 1273.
6	倪世宏, 史忠科, 谢川, 等. 军用战机机动飞行动作识别知识库的建立[J]. 计算机仿真, 2005, 22 (4): 23- 26. doi: 10.3969/j.issn.1006-9348.2005.04.007
	NI S H , SHI Z K , XIE C , et al. Establishment of knowledge base for recognition of manimaneuver flight action of military fighter[J]. Computer Simulation, 2005, 22 (4): 23- 26. doi: 10.3969/j.issn.1006-9348.2005.04.007
7	WANG J , XIONG B , YAN M O . Recognition method of helicopter flight condition based onrandom forest[J]. Computer Engineering and Applications, 2017, 53 (17): 149- 152, 159.
8	CHEN X, ZHANG X G, WANG W J, et al. Multi-objective Monte-Carlo tree search based aerial maneuvering control[C]//Proc. of the IEEE Chinese Guidance, Navigation and Control Conference, 2016: 81-87.
9	RODIN E Y , AMIN M . Maneuver prediction in air combat via artificial neural networks[J]. Computers & Mathematics with Applications, 1992, 24 (3): 95- 112.
10	李丹丹, 路辉, 郎荣玲. 基于主成分分析和SVM的飞参阶段划分研究[J]. 现代电子技术, 2010, 33 (8): 134- 137. doi: 10.3969/j.issn.1004-373X.2010.08.042
	LI D D , LU H , LANG R L . Research on stage division of flying parameter based on principal component analysis and SVM[J]. Modern Electronic Technology, 2010, 33 (8): 134- 137. doi: 10.3969/j.issn.1004-373X.2010.08.042
11	HUANG C Q , DONG K S , HUANG H Q , et al. Autonomous air combat maneuver decision using Bayesian inference and moving horizon optimization[J]. Journal of Systems Engineering and Electronics, 2018, 29 (1): 86- 97. doi: 10.21629/JSEE.2018.01.09
12	XU X M , YANG R N , FU Y . Situation assessment for air combat based on novel semi-supervised naive Bayes[J]. Journal of Systems Engineering and Electronics, 2018, 29 (4): 768- 779. doi: 10.21629/JSEE.2018.04.11
13	CHUA L O , ROSKA T . The CNN paradigm[J]. IEEE Trans.on Circuits & Systems I: Fundamental Theory & Applications, 1993, 40 (3): 147- 156.
14	SANDHYA R S , APPARAO N G , USHA S V . Kinematic joint descriptor and depth motion descriptor with convolutional neural networks for human action recognition[J]. Materials Today: Proceedings, 2021, 37 (2): 3164- 3173.
15	JIAN Q Z , XIAO L G . Deep convolution neural networks for twitter sentiment analysis[J]. IEEE Access, 2018, 6, 23253- 23260. doi: 10.1109/ACCESS.2017.2776930
16	REN S Q , HE K M , GIRSHICK R , et al. Faster R-CNN: towards real-time object detection with region proposal networks[J]. IEEE Trans.on Pattern Analysis & Machine Intelligence, 2017, 39 (6): 1137- 1149.
17	HOCHREITER S , SCHMIDHUBER J . Long short-term memory[J]. Neural Computation, 1997, 9 (8): 1735- 1780. doi: 10.1162/neco.1997.9.8.1735
18	HUANG M, TAO W, LIANG X, et al. Application of improved long-short-term memory network in human morphology detection[C]//Proc. of the IEEE 7th International Conference on Computer Science and Network Technology, 2019: 111-114.
19	BIN Y , YANG Y , SHEN F M , et al. Describing video with attention-based bidirectional LSTM[J]. IEEE Trans.on Cybernetics, 2019, 49 (7): 2631- 2641. doi: 10.1109/TCYB.2018.2831447
20	WU C , CHEN L , WANG G B , et al. Spatiotemporal scenario generation of traffic flow based on LSTM-GAN[J]. IEEE Access, 2020, 8, 186191- 186198. doi: 10.1109/ACCESS.2020.3029230
21	MA L , TIAN S . A hybrid CNN-LSTM model for aircraft 4D tra-jectory prediction[J]. IEEE Access, 2020, 8, 134668- 134680. doi: 10.1109/ACCESS.2020.3010963
22	WU J L , HE Y , YU L C , et al. Identifying emotion labels from psychiatric social texts using a bi-directional LSTM-CNN model[J]. IEEE Access, 2020, 8, 66638- 66646. doi: 10.1109/ACCESS.2020.2985228
23	BELLAMY J C . Airborne navigational recorders[J]. Navigation Journal of the Institute of Navigation, 1952, 3 (6): 182- 188.
24	ROBERTS C A . The status of flight recorders in modern aircraft[J]. Journal of Air Law & Commerce, 1977, 43 (2): 271- 287.
25	史浩宇, 赵克新. 无人作战飞机自主空战战术动作的分层识别方法[J]. 火力与指挥控制, 2019, 44 (12): 51- 55, 60. doi: 10.3969/j.issn.1002-0640.2019.12.011
	SHI H Y , ZHAO K X . Hierarchical identification method for autonomous air combat tactical movements of unmanned combat aircraft[J]. Fire Control & Command Control, 2019, 44 (12): 51- 55, 60. doi: 10.3969/j.issn.1002-0640.2019.12.011
26	AUSTIN F , CARBONE G , HINZ H , et al. Game theory for automated maneuvering during air-to-air combat[J]. Journal of Guidance Control & Dynamics, 1990, 13 (6): 1143- 1149.
27	彭阳, 彭晓明. 一种无人机飞行动作质量分析方法[J]. 飞行力学, 2017, 35 (6): 65- 69. doi: 10.3969/j.issn.1002-0853.2017.06.015
	PENG Y , PENG X M . A method for quality analysis of UAV flight action[J]. Journal of Flight Mechanics, 2017, 35 (6): 65- 69. doi: 10.3969/j.issn.1002-0853.2017.06.015
28	方伟, 王玉, 闫文君, 等. 基于微分思想和卷积神经网络的飞行动作识别[J]. 中国电子科学研究院学报, 2021, 16 (4): 347- 353. doi: 10.3969/j.issn.1673-5692.2021.04.008
	FANG W , WANG Y , YAN W J , et al. Flight action recognition based on differential idea and convolutional neural network[J]. Journal of China Academy of Electronic Sciences, 2021, 16 (4): 347- 353. doi: 10.3969/j.issn.1673-5692.2021.04.008
29	XIE C , KUMAR A . Finger vein identification using convolutional neural network and supervised discrete hashing[J]. Pattern Recognition Letters, 2019, 119 (3): 148- 156.
30	VENKATA R M , ARUNKUMAR R . A weight based feature extraction model on multifaceted multimedia bigdata using convolutional neural network[J]. International Information and Engineering Technology Association, 2020, 25 (6): 725- 729.
31	XU K , GONG H , LIU F . Vehicle detection based on improved multitask cascaded convolutional neural network and mixed image enhancement[J]. IET Image Processing, 2020, 14 (17): 4621- 4632. doi: 10.1049/iet-ipr.2020.1005
32	SHI Q Q , LI W , TAO R , et al. Ship classification based on multifeature ensemble with convolutional neural network[J]. Remote Sensing, 2019, 11 (4): 419- 440. doi: 10.3390/rs11040419
33	GRAVES A , SCHMIDHUBER J . Framewise phoneme classification with bidirectional LSTM and other neural network architectures[J]. Neural Networks, 2005, 18 (6): 602- 610.
34	孟光磊, 陈振, 罗元强. 基于动态贝叶斯网络的机动动作识别方法[J]. 系统仿真学报, 2017, 29 (S1): 140- 145.
	MENG G L , CHEN Z , LUO Y Q . Maneuvering action recognition method based on dynamic Bayesian network[J]. Journal of System Simulation, 2017, 29 (S1): 140- 145.
35	李鸿利, 单征, 郭浩然. 基于MDTW的飞行动作识别算法[J]. 计算机工程与应用, 2015, 51 (9): 267- 270. doi: 10.3778/j.issn.1002-8331.1307-0018
	LI H L , SHAN Z , GUO H R . Flight action recognition algorithm based on MDTW[J]. Computer Engineering and Applications, 2015, 51 (9): 267- 270. doi: 10.3778/j.issn.1002-8331.1307-0018

序号	基础类	基本飞行动作	高度	航向角	俯仰角	滚转角	滚转率	转弯率
1	平飞类	直线上升飞行	保持	保持	正保持	零保持	零保持	零保持
		直线水平飞行	增加	保持	零保持	零保持	零保持	零保持
		直线下降飞行	减小	保持	负保持	零保持	零保持	零保持
2	转弯类	右转水平飞行	保持	正变化	零保持	零保持	零保持	变化
		左转水平飞行	保持	负变化	零保持	零保持	零保持	变化
		右转上升飞行	增加	正变化	正变化	零保持	零保持	变化
		右转下降飞行	减小	正变化	负变化	零保持	零保持	变化
		左转上升飞行	增加	负变化	正变化	零保持	零保持	变化
		左转下降飞行	减小	负变化	负变化	零保持	零保持	变化
		垂直上升转弯飞行	增加	零保持	正变化	零保持	零保持	变化
		垂直下降转弯飞行	减小	零保持	负变化	零保持	零保持	变化
3	滚转类	直线滚转飞行	保持	零保持	零保持	变化	变化	零保持
3	滚转类	斜滚转飞行	变化	变化	变化	变化	变化	零保持

序号	复杂飞行动作	所包含的基本飞行动作
1	告警改出	直线水平、垂直上升转弯、直线上升、斜横滚、飞行动作调整段、垂直下降转弯、斜横滚、飞行动作调整段、直线水平
2	筋斗	直线水平、垂直上升转弯、直下降转弯、飞行动作调整段、直线水平
3	大姿态回转	直线水平、右转/左转上升、斜横滚、飞行动作调整段、直线水平
4	殷麦曼	直线水平、垂直上升、直线滚转、飞行动作调整段、直线水平
5	半滚倒转	直线水平、直线滚转、垂直上升、飞行动作调整段、直线水平

符号化表示	基本飞行动作	符号化表示	基本飞行动作
0	直线上升飞行	7	左转上升飞行
1	直线水平飞行	8	左转下降飞行
2	直线下降飞行	9	垂直上升转弯飞行
3	右转水平飞行	10	直线滚转飞行
4	左转水平飞行	11	斜滚转飞行
5	右转上升飞行	12	垂直下降转弯飞行
6	右转下降飞行	-	-

层	输出尺寸	参数
填充层_1	(None, 1, 8, 14)	0
卷积层_1	(None, 256, 7, 13)	1 280
丢弃层_1	(None, 256, 7, 13)	0
填充层_2	(None, 256, 7, 17)	0
卷积层_2	(None, 80, 6, 16)	82 000
丢弃层_2	(None, 80, 6, 16)	0
展平层	(None, 7 680)	0
全连接层_1	(None, 256)	1 966 336
丢弃层_3	(None, 256)	0
全连接层_2	(None, 13)	3 341

层	输出尺寸	参数
嵌入层	(None, 115, 16)	208
LSTM层	(None, 16)	2 112
丢弃层	(None, 16)	0
全连接层	(None, 2)	34