面向低资源的无人机指令意图识别算法及半实物仿真

doi:10.16182/j.issn1004731x.joss.24-0566

摘要/Abstract

摘要：

在通信网络部分失能或被干扰时，无人机陷入“低资源环境”，必须依赖本地硬件资源，面临着计算能力、存储空间和能源供应的限制。针对“低资源环境”下的无人机指令意图识别研究需求，设计并实现了一个应急救灾场景中无人机指令意图识别半实物仿真系统。基于“低资源环境”的机载硬件在环，通过GIS+BIM三维环境建模任务场景，半实物仿真无人机指令意图识别与任务规划。针对核心功能指令意图识别提出了一种新的轻量化算法，基于GraphSAGE的全局句子结构信息抽取与FastText局部语义特征的共同注意力融合机制，优化提升了意图理解预测的准确率和响应速度。在构建的专业无人机指令意图数据集上，半实物仿真验证指令意图识别准确率为0.890 7、时间为58.808 ms，满足实时性要求。

关键词: 无人机, 低资源环境, 指令意图识别, 文本分类, 轻量化算法, 半实物仿真

Abstract:

When a communication network is partially disabled or disrupted, an UAV is plunged into a "low-resource environment" and must rely on local hardware resources. This situation imposes constraints on computing power, storage capacity, and energy availability. To address the need for command intent recognition in such environments, a semi-physical simulation system for UAV in emergency rescue operations has been designed and implemented. Based on the low resource airborne hardware in the loop, the system simulates UAV command intention recognition and mission planning through GIS+BIM 3D environment modeling task scenarios. A new lightweight algorithm for intent recognition has been proposed, based on the joint attention fusion mechanism of global sentence structure information extraction using GraphSAGE and local semantic features of FastText, which optimizes and improves the accuracy and response speed of intent understanding prediction. On the constructed professional UAV command intent dataset, semi-physical simulation verifies that the accuracy of command intention recognition is 0.890 7 and the time is 58.808 ms, which meets the real-time requirement.

Key words: UAV, low-resource environments, command intent recognition, text classification, lightweight algorithm, semi-physical simulation

中图分类号:

TP391

刘鸿福,付雅晶,张万鹏等 . 面向低资源的无人机指令意图识别算法及半实物仿真[J]. 系统仿真学报, 2024, 36(12): 2894-2905.

Liu Hongfu,Fu Yajing,Zhang Wanpeng,et al . Algorithm and Semi-physical System Simulation for Command Intent Recognition of UAV in Low-resource Environment[J]. Journal of System Simulation, 2024, 36(12): 2894-2905.

图/表 12

图1

图2

图3

图4

图5

图6

图7

图8

图9

表1

表2

表3

参考文献 22

1	Duan Haibin, Shao Shan, Su Bingwei, et al. New Development Thoughts on the Bio-inspired Intelligence Based Control for Unmanned Combat Aerial Vehicle[J]. Science China Technological Sciences, 2010, 53(8): 2025-2031.
2	王正义, 陈邓安. 美无人机系统当前和今后担负的任务及其发展趋势[J]. 中国电子科学研究院学报, 2023, 18(2): 195-202.
	Wang Zhengyi, Chen Dengan. The Present and Future Tasks and Development Trends of Unmanned Aircraft System in the United States[J]. Journal of China Academy of Electronics and Information Technology, 2023, 18(2): 195-202.
3	Wang Yuhao, Pang Yutian, Gorceski Stojanche, et al. A Voice Communication-augmented Simulation Framework for Aircraft Trajectory Simulation[J]. IEEE Transactions on Intelligent Transportation Systems, 2022, 23(7): 7310-7320.
4	Aletras N, Tsarapatsanis D, Preoţiuc-Pietro D, et al. Predicting Judicial Decisions of the European Court of Human Rights: A Natural Language Processing Perspective[J]. PeerJ Computer Science, 2016, 2: e93.
5	王康静, 钱江海. 一种融合改进TF-IDF与词典模型的情感分类算法[J]. 上海电力大学学报, 2024, 40(1): 80-86.
	Wang Kangjing, Qian Jianghai. An Improved Emotion Classification Algorithm Based on Improved TF-IDF and Dictionary Model[J]. Journal of Shanghai University of Electric Power, 2024, 40(1): 80-86.
6	Jeong Minwoo, Geunbae Lee Gary. Jointly Predicting Dialog Act and Named Entity for Spoken Language Understanding[C]//2006 IEEE Spoken Language Technology Workshop. Piscataway: IEEE, 2006: 66-69.
7	Zhang Chenwei, Li Yaliang, Du Nan, et al. Joint Slot Filling and Intent Detection via Capsule Neural Networks[C]//Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics. Stroudsburg: ACL, 2019: 5259-5267.
8	刘鹏宇, 朱雪耀. 基于深度学习的融合空域空管指令语义解析技术[J]. 航空学报, 2023, 44(增1): 155-163.
	Liu Pengyu, Zhu Xueyao. Semantic Parsing Technology of Air Traffic Control Instruction in Fusion Airspace Based on Deep Learning[J]. Acta Aeronautica et Astronautica Sinica, 2023, 44(S1): 155-163.
9	符凯, 朱雪耀, 吕全喜, 等. 基于深度学习的无人机指令意图识别技术[J]. 兵工自动化, 2022, 41(10): 41-44, 59.
	Fu Kai, Zhu Xueyao, Quanxi Lü, et al. UAV Command Intent Identification Technology Based on Deep Learning[J]. Ordnance Industry Automation, 2022, 41(10): 41-44, 59.
10	关景文, 宋晓, 李晓庆, 等. 导弹领域文本嵌套命名实体识别方法研究[J]. 系统仿真学报, 2023, 35(8): 1757-1767.
	Guan Jingwen, Song Xiao, Li Xiaoqing, et al. Research on Nested Named Entity Recognition in Missile Field Text[J]. Journal of System Simulation, 2023, 35(8): 1757-1767.
11	Zhang Pan, Huang Li. Research on Chinese Intent Recognition Based on BERT Pre-trained Model[C]//Proceedings of the 2020 5th International Conference on Mathematics and Artificial Intelligence. New York: ACM, 2020: 128-132.
12	Wang Yaqing, Wang Song, Li Yanyan, et al. Recognizing Medical Search Query Intent by Few-shot Learning[C]//Proceedings of the 45th International ACM SIGIR Conference on Research and Development in Information Retrieval. New York: ACM, 2022: 502-512.
13	Sharma Bidisha, Madhavi Maulik, Li Haizhou. Leveraging Acoustic and Linguistic Embeddings from Pretrained Speech and Language Models for Intent Classification[C]//ICASSP 2021-2021 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). Piscataway: IEEE, 2021: 7498-7502.
15	Hamilton W L, Ying R, Leskovec J. Inductive Representation Learning on Large Graphs[C]//Proceedings of the 31st International Conference on Neural Information Processing Systems. Red Hook: Curran Associates Inc., 2017: 1025-1035.
16	Bojanowski P, Grave E, Joulin A, et al. Enriching Word Vectors with Subword Information[J]. Transactions of the Association for Computational Linguistics, 2017, 5: 135-146.
17	Chen Y. Convolutional Neural Network for Sentence Classification[D]. Waterloo: University of Waterloo, 2015.
18	Liu Pengfei, Qiu Xipeng, Huang Xuanjing. Recurrent Neural Network for Text Classification with Multi-task Learning[C]//Proceedings of the Twenty-Fifth International Joint Conference on Artificial Intelligence. Palo Alto: AAAI Press, 2016: 2873-2879.
19	Lai Siwei, Xu Liheng, Liu Kang, et al. Recurrent Convolutional Neural Networks for Text Classification[C]//Proceedings of the Twenty-ninth AAAI Conference on Artificial Intelligence. Palo Alto: AAAI Press, 2015: 2267-2273.
20	Johnson R, Zhang Tong. Deep Pyramid Convolutional Neural Networks for Text Categorization[C]//Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics. Stroudsburg: ACL, 2017: 562-570.
21	Devlin J, Chang Mingwei, Lee K, et al. BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding[C]//Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies. Stroudsburg: ACL, 2019: 4171-4186.
22	Sun Yu, Wang Shuohuan, Li Yukun, et al. ERNIE: Enhanced Representation Through Knowledge Integration[EB/OL]. (2019-04-19) [2024-04-20]. .

算法	A	F₁	R
TextCNN	0.875 2	0.873 9	0.875 0
TextRNN	0.863 5	0.855 6	0.862 1
TextRCNN	0.875 4	0.875 3	0.874 8
DPCNN	0.872 5	0.872 0	0.870 0
BERT	0.897 2	0.888 2	0.887 0
ERNIE	0.881 0	0.877 7	0.874 3
本文	0.890 7	0.889 0	0.889 0

算法	A	F₁	R
本文	0.890 7	0.889 0	0.889 0
Ⅰ	0.673 8	0.680 3	0.680 0
Ⅱ	0.867 2	0.863 5	0.863 5
Ⅲ	0.881 5	0.878 6	0.880 0
Ⅳ	0.793 4	0.800 7	0.793 0

算法	训练时间/min	预测时间/ms
TextCNN	21.317	507.239
TextRNN	43.617	1 042.942
TextRCNN	26.100	684.258
DPCNN	33.217	671.357
BERT	158.233	13 608.010
ERNIE	187.875	13 151.264
本文模型	29.633	58.808

[1]	李二超, 张生辉. 基于DMOEA-APTC算法的无人机在线航迹规划[J]. 系统仿真学报, 2024, 36(9): 2086-2099.
[2]	刘斌, 兰莹, 黄文焘, 范勤勤. 融合动态窗口法的无人机动态路径规划算法[J]. 系统仿真学报, 2024, 36(8): 1843-1853.
[3]	黄志锋, 刘媛华. 基于改进哈里斯鹰和B-spline曲线的无人机路径规划研究[J]. 系统仿真学报, 2024, 36(7): 1509-1524.
[4]	王涛, 吉晓东. 混合概率信道下的全双工无人机中继轨迹优化[J]. 系统仿真学报, 2024, 36(6): 1369-1377.
[5]	李勇波, 田润梅, 张辉, 郭善鹏, 李琪. 基于Windows/RTX的实时仿测软件设计[J]. 系统仿真学报, 2024, 36(6): 1468-1474.
[6]	高超, 黄郑, 赵轩, 王红星, 龙涛. 基于构型优选的5G集群无人机协同导航方法[J]. 系统仿真学报, 2024, 36(4): 981-990.
[7]	余翔, 邓千锐, 段思睿, 姜陈. 一种多无人机协同优先覆盖搜索算法[J]. 系统仿真学报, 2024, 36(4): 991-1000.
[8]	桂雪琪, 李春涛. 基于视野和速度引导的无人机集群避障算法[J]. 系统仿真学报, 2024, 36(3): 545-554.
[9]	李高扬, 黎向锋, 赵康, 金玉超, 易志东, 左敦稳. 全粒子推动野马优化算法的无人机三维路径规划[J]. 系统仿真学报, 2024, 36(3): 595-607.
[10]	卢元杰, 龙珊珊, 赵航, 冯国旭, 赵晓葭. 基于混合模型的异构无人机蜂群效能评估[J]. 系统仿真学报, 2024, 36(3): 700-712.
[11]	马赟, 王源, 李蒙, 王鹏, 汤艳玲. 非平面全驱动多旋翼无人机的鲁棒预测控制[J]. 系统仿真学报, 2024, 36(2): 415-422.
[12]	豆建斌, 王小兵, 杨红坚, 高玉龙. 成像制导导弹试验鉴定半实物仿真系统设计与应用[J]. 系统仿真学报, 2024, 36(2): 522-532.
[13]	石成泷, 华翔, 王东, 张金金, 蒋天启, 党元章. 考虑局部特征的无人机集群关键节点识别方法[J]. 系统仿真学报, 2024, 36(12): 2824-2833.
[14]	王鹏, 刘昊雨, 李妮, 于泽熙, 贾尚杰. 基于改进ABC算法的有-无人协同空战行为建模[J]. 系统仿真学报, 2024, 36(12): 2871-2883.
[15]	杨哲, 崔颖函, 郭灵犀, 李嘉鑫, 吴旭生. 融合数据仿真与深度学习算法的飞行器残骸搜寻技术[J]. 系统仿真学报, 2024, 36(10): 2238-2245.