多域联合表征的电网无线通信链路质量预测

Multi-domain joint representation-based wireless communication link quality prediction for power grids

摘要

针对智能电网无线通信链路质量动态预测中存在的动态表征能力不足、预测精度低等问题，提出一种基于多域联合表征的链路质量预测框架。首先，设计 CNN-Transformer 混合时域编码器，结合卷积神经网络的局部特征提取能力与自注意力机制的全局时序建模优势，精准捕获序列的非线性波动规律。其次，采用基于复值深度可分离卷积的时频编码器，直接处理时频变换生成的复数谱，完整保留幅度−相位关联特性。最后，提出对称式多域特征交叉融合模块，通过双向特征交互实现时域与时频域的语义对齐，并引入对比损失函数增强多域特征的可判别性。实验基于连续 48 h 采集的真实电网无线通信链路数据进行验证，结果表明，所提方法在链路质量预测任务中较主流单域模型最高提升 7.3%，验证了多域联合表征的有效性。

Abstract

To address the challenges of insufficient dynamic representation and low prediction accuracy in wireless link quality prediction for smart grids, this paper proposes a multi-domain joint representation framework. First, a CNN (convolutional neural network)-Transformer hybrid temporal encoder is designed to integrate the local feature extraction capability of CNNs with the global temporal modeling advantages of self-attention mechanisms, accurately capturing nonlinear fluctuation patterns in sequences. Second, a complex-valued depthwise separable convolution-based time-frequency encoder is developed to directly process the complex spectrogram generated by time-frequency transformation, preserving amplitude-phase correlation characteristics. Finally, a symmetric multi-domain feature cross-fusion module is proposed to achieve semantic alignment between temporal and time-frequency domains through bidirectional feature interaction, with a contrastive loss function introduced to enhance feature discriminability. Experiments on 48-hour continuous real-world grid wireless link data demonstrate that the proposed method achieves a maximum improvement of 7.3% in link quality prediction tasks compared to mainstream single-domain models, validating the effectiveness of multi-domain joint representation.

关键词

无线通信链路质量预测 / 多域联合表征 / 复值卷积网络 / 跨域融合

Key words

wireless communication link quality prediction / multi-domain joint representation / complex-valued convolutional network / cross-domain fusion

参考文献

[1] Krizhevsky A, Sutskever I, Hinton G E. ImageNet classification with deep convolutional neural networks[J].Communications of the ACM, 2017, 60(6): 84-90.

[2] Iandola F N, Han Song, Moskewicz M W, et al. SqueezeNet: AlexNet-level accuracy with 50x fewer parameters and <0.5 MB model size[PP/OL]. V4.arXiv (2016-11-04)[2024-10-11].https://arxiv.org/abs/1602.07360.

[3] Vaswani A, Shazeer N, Parmar N, et al. Attention is all you need[C]//Proceedings of the 31st International Conference on Neural Information Processing Systems. Long Beach: Curran Associates Inc., 2017: 6000-6010.

[4] 刘杰, 金勇杰, 田明 . 基于 VMD 和 TCN 的多尺度短期电力负荷预测[J].电子科技大学学报, 2022, 51(4): 550-557.

[5] Liu Jie, Jin Yongjie, Tian Ming . Multi-scale short-term load forecasting based on VMD and TCN[J].Journal of University of Electronic Science and Technology of China, 2022, 51(4): 550-557. (in Chinese)

[6] 黄颖, 许剑, 周子祺, 等. 高效长序列水位预测模型的研究与实现[J].电子科技大学学报, 2023, 52(4): 595-601.

[7] Huang Ying, Xu Jian, Zhou Ziqi, et al. Research and implementation of efficient long sequence model for water level forecasting[J].Journal of University of Electronic Science and Technology of China, 2023, 52(4): 595-601. (in Chinese)

[8] Xu Ming, Liu Wei, Xu Jinwei, et al. Recurrent neural network based link quality prediction for fluctuating low power wireless links[J].Sensors, 2022, 22(3): 1212.

[9] Zha Minghu, Zhu Li, Zhu Yunyun, et al. A novel SCNN-LSTM model for predicting the SNR confidence interval in wearable wireless sensor network[J].Intelligent Systems with Applications, 2024, 22: 200363.

[10] 向征 . 基于 LSTM 的智能电网链路质量置信区间预测[J].电测与仪表, 2021, 58(11): 93-100.

[11] Xiang Zheng . Link quality confidence interval prediction of smart grid based on LSTM[J].Electrical Measurement & Instrumentation, 2021, 58(11): 93-100. (in Chinese)

[12] 孙伟, 李鹏宇, 杨建平, 等. 配电泛在物联网无线通信链路可靠性的置信区间预测[J].电子测量与仪器学报, 2020, 34(6): 32-40.

[13] Sun Wei, Li Pengyu, Yang Jianping, et al. Reliability confidence interval prediction of power distribution ubiquitous IoT wireless communication link[J].Journal of Electronic Measurement and Instrumentation, 2020, 34(6): 32-40. (in Chinese)

[14] Xue Xue, Sun Wei, Wang Jianping, et al. RVFL-LQP: RVFL-based link quality prediction of wireless sensor networks in smart grid[J].IEEE Access, 2020, 8: 7829-7841.

[15] Liu Linlan, Niu Mingxiao, Zhang Chao, et al. Light gradient boosting machine-based link quality prediction for wireless sensor networks[J].Wireless Communications and Mobile Computing, 2022, 2022(1): 8278087.

[16] Feng Yi, Liu Linlan, Shu Jian. A link quality prediction method for wireless sensor networks based on XGBoost[J].IEEE Access, 2019, 7: 155229-155241.

[17] 刘琳岚, 肖庭忠, 舒坚, 等. 基于门控循环单元的链路质量预测[J].工程科学与技术, 2022, 54(6): 51-58.

[18] Liu Linlan, Xiao Tingzhong, Shu Jian, et al. Link quality prediction based on gate recurrent unit[J].Advanced Engineering Sciences, 2022, 54(6): 51-58. (in Chinese)

[19] Kim D, Nam H, Kim D. Adaptive code dissemination based on link quality in wireless sensor networks[J].IEEE Internet of Things Journal, 2017, 4(3): 685-695.

[20] Lowrance C J, Lauf A P. Link quality estimation in ad hoc and mesh networks: A survey and future directions[J].Wireless Personal Communications, 2017, 96(1): 475-508.

[21] He Kaiming, Zhang Xiangyu, Ren Shaoqing, et al. Deep residual learning for image recognition[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. Las Vegas: IEEE, 2016: 770-778.

[22] Trabelsi C, Bilaniuk O, Zhang Ying, et al. Deep complex networks[C]//Proceedings of the 6th International Conference on Learning Representations. Vancouver: OpenReview.net, 2018.

[23] Howard A G, Zhu Menglong, Chen Bo, et al. MobileNets: Efficient convolutional neural networks for mobile vision applications[PP/OL]. V1.arXiv (2017-04-17)[2024-11-25].https://arxiv.org/abs/1704.04861.

[24] Dosovitskiy A, Beyer L, Kolesnikov A, et al. An image is worth 16x16 words: Transformers for image recognition at scale[C]//Proceedings of the 9th International Conference on Learning Representations. [S.l.]: OpenReview.net,2021.

[25] Tan Mingxing, Le Q V. EfficientNetV2: Smaller models and faster training[C]//Proceedings of the 38th International Conference on Machine Learning (ICML). [S.l.]: ICML,2021: 10096-10106.

基金资助

国网四川省电力公司科技项目(5210012300AJ)