基于XGBoost算法的水下无线电能传输系统多参数快速辨识方法

doi:10.19595/j.cnki.1000-6753.tces.252090

摘要
图/表
参考文献
相关文章 (15)

全文: PDF (2516 KB)
输出: BibTeX | EndNote (RIS)

摘要

耦合机构的参数辨识是保证无线电能传输系统（WPT）高效传输与稳定控制的前提。然而,由于海水高电导率特性,水下磁耦合谐振式无线电能传输（MCR-WPT）系统在电能传输过程中会产生涡流损耗,相比于空气环境其磁耦合机构模型参数更为复杂多元,导致传统基于空气环境下的参数辨识技术难以满足其耦合机构多元参数的辨识需求。因此,本文提出了一种基于数据驱动的水下WPT系统多参数在线辨识策略,通过预先采集耦合参数与接收端Buck输入电压的数据集,并使用XGBoost算法拟合其非线性映射函数,便可在实际工作中通过测量接收端Buck电路的直流输入电压实现水下WPT系统耦合参数的在线快速辨识,既避免了水下双端通信延迟干扰,也不需要高频采样模块,且更快的辨识速度能够更好地适应海水频繁冲击扰动的海洋环境,为系统实时控制提供支撑。此外,XGBoost算法相比于传统人工神经网络（ANN）算法所需数据量更少,泛化能力更好,保证了系统在不同电导率海水中的可靠应用。本文在不同线圈偏移条件下对比了XGBoost算法与传统ANN算法的辨识性能,证明在同数量样本训练中,XGBoost算法辨识精度与辨识速度均优于ANN算法。同时,本文在不同电导率的海洋环境下验证了XGBoost算法的泛化能力。实验结果表明,本文所使用的XGBoost算法在不同偏移条件、不同电导率海洋环境下对水下WPT系统耦合参数的辨识误差均小于4%,平均辨识误差为1.93%,平均计算时间约为10ms,能够准确快速地实现耦合参数的在线辨识。

	服务

	把本文推荐给朋友
	加入我的书架
	加入引用管理器
	E-mail Alert
	RSS
	作者相关文章
	罗博
	曾雪梅
	吴欢
	白龙雷
	游江

关键词 ：水下无人航行器, 水下无线电能传输系统, 参数辨识技术, XGBoost算法

Abstract：

A continuous and reliable power supply is essential for ensuring the long-term, efficient operation of Unmanned Underwater Vehicles (UUVs). Currently, the main methods for UUV power supply include battery replacement and cable power supply, both of which have drawbacks such as low safety, poor concealment, and low automation levels. Magnetically-Coupled Resonant Wireless Power Transfer (MCR-WPT) technology uses electromagnetic fields as the medium for power transmission, eliminating the need for plug-in connections. This significantly enhances the safety, reliability, and automation of underwater equipment power supply, providing a new solution for UUV power replenishment. The parameter identification of the coupling mechanism is crucial for the efficient power transmission and stable control of the Wireless Power Transfer (WPT) system. However, due to the high conductivity of seawater, the MCR-WPT system in underwater environments generates eddy current losses during power transmission. Compared to air environments, the magnetic coupling mechanism's model parameters in underwater conditions are more complex and varied, making traditional parameter identification techniques based on air environments insufficient for identifying the multiple parameters of the coupling mechanism.
Therefore, this paper proposes a data-driven multi-parameter online identification strategy for underwater WPT systems. By pre-collecting data sets of coupling parameters and the input voltage of the receiver’s Buck circuit, and using the XGBoost algorithm to fit the nonlinear mapping function, the online and rapid identification of underwater WPT system coupling parameters can be achieved by measuring the DC input voltage of the receiver's Buck circuit during actual operation. This method not only effectively solves the problem of identifying multiple parasitic parameters caused by eddy current losses in underwater environments, but also provides a faster identification speed that better adapts to frequent disturbances caused by ocean currents, enabling quicker model parameter references for subsequent system control. Moreover, the proposed method only requires the collection of the DC voltage at the receiver end of the wireless power transmission system to achieve online parameter identification. This avoids interference from underwater bilateral communication delays and eliminates the need for high-frequency sampling modules, significantly saving hardware resources and improving system reliability. Additionally, the XGBoost algorithm used in this paper requires less data compared to traditional Artificial Neural Networks (ANN) algorithms, is more streamlined, and demonstrates better generalization capabilities, ensuring the system’s reliable application in seawater with varying conductivity.
Finally, an experimental platform for parameter identification of the underwater WPT system based on the XGBoost algorithm was built. The performance of the XGBoost algorithm was compared with the traditional ANN algorithm under different coil offset conditions. The results demonstrate that, with the same number of samples, the XGBoost algorithm outperforms the ANN algorithm in both identification accuracy and speed. Furthermore, the generalization ability of the XGBoost algorithm was validated in seawater environments with varying conductivity. Experimental results show that the XGBoost algorithm achieves a coupling parameter identification error of less than 4% under different offset conditions and seawater conductivity environments, with an average identification error of 1.93% and an average computation time of approximately 10 ms. This confirms that the XGBoost algorithm can accurately and quickly identify coupling parameters online.

Key words： Unmanned Underwater Vehicles Underwater Wireless Power Transfer System Parameter Identification Technology XGBoost Algorithm

收稿日期: 2025-12-10

PACS:

TM724

基金资助:

哈尔滨工程大学智能海洋航行器技术全国重点实验室基金（2026-HYHXQ-WDZC-06）、海南省自然科学基金（526MS0258）、国家自然科学基金青年基金（52307006）、烟台市科技创新发展计划基础研究类项目（2024JCYJ084）资助

作者简介: 罗博男,1992年生,副教授,博士生导师,研究方向为无线电能传输技术。E-mail：boluo@hrbeu.edu.cn;曾雪梅女,2001年生,硕士研究生,研究方向为水下无线电能传输系统参数辨识与先进控制技术。E-mail：324517003@hrbeu.edu.cn

引用本文:

罗博, 曾雪梅, 吴欢, 白龙雷, 游江. 基于XGBoost算法的水下无线电能传输系统多参数快速辨识方法[J]. 电工技术学报, 0, (): 38-. Luo Bo, Zeng Xuemei, Wu Huan, Bai Longlei, You Jiang. A Fast Multi-Parameter Identification Method for Underwater Wireless Power Transfer System Based on XGBoost Algorithm. Transactions of China Electrotechnical Society, 0, (): 38-.

链接本文:

https://dgjsxb.ces-transaction.com/CN/10.19595/j.cnki.1000-6753.tces.252090 https://dgjsxb.ces-transaction.com/CN/Y0/V/I/38

[1] 招聪,谢仁和,徐纪伟,等.深海能源基站技术现状与发展趋势[J]. 舰船科学技术,2023,45(16):7-12.
Zhao Cong, Xie Renhe, Xu Jiwei, et al.Status and development trends of deep sea energy base station technology[J]. Ship Science and Technology, 2023, 45(16): 7-12.
[2] Zhang Renjie, Yuan Huan, Rong Mingzhe, et al.Self-Tuning WPT System with Constant Voltage Output Under Resonance Frequency Shift[J]. IEEE Transactions on Power Electronics, 2024, 39(01): 1713-1722.
[3] 丰利军,朱春波,张剑韬,等.水下无人航行器水下无线充电关键技术研究[J].舰船科学技术, 2020, 42(23): 159-162.
Feng Lijun, Zhu Chunbo, Zhang Jiantao, et al.Research on key technology based on wireless charging technology for unmanned underwater vehicle[J]. Ship Science and Technology, 2020, 42(23): 159-162.
[4] 薛明,杨庆新,章鹏程,等.无线电能传输技术应用研究现状与关键问题[J].电工技术学报, 2021, 36(08): 1547-1568.
Xue Ming, Yang Qingxin, Zhang Pengcheng, et al.Application Status and Key Issues of Wireless Power Transmission Technology[J]. Transactions of China Electrotechnical Society, 2021, 36(08): 1547-1568.
[5] 闫争超,胡谦宇,赵晨旭,等.水下航行器感应式无线电能传输技术研究综述[J].中国电机工程学报, 2023, 43(24): 9668-9682.
Yan Zhengchao, Hu Qianyu, Zhao Chenxu, et al.Review on Inductive Wireless Power Transfer Technology for Underwater Vehicles[J]. Proceedings of the CSEE, 2023, 43(24): 9668-9682.
[6] 吴旭升,孙盼,杨深钦,等.水下无线电能传输技术及应用研究综述[J].电工技术学报, 2019, 34(08): 1559-1568.
Wu Xusheng, Sun Pan, Yang Shenqin, et al.Review on Underwater Wireless Power Transfer Technology and Its Application[J]. Transactions of China Electrotechnical Society, 2019, 34(08): 1559-1568.
[7] Dai Xin, Li Xiaofei, Li Yanling, et al.Maximum Efficiency Tracking for Wireless Power Transfer Systems with Dynamic Coupling Coefficient Estimation[J]. IEEE Transactions on Power Electronics, 2018, 33(06): 5005-5015.
[8] 张岩,都义彬,臧朝辉,等.基于发射端信息的无线电能传输系统负载与互感辨识方法研究[J/OL]. 电工技术学报, 1-11.
Zhang Yan, Du Yibin, Zang Chaohui, et al.Load and Mutual Inductance Identification of WPT System Based on Transmitter Information[J/OL]. Transactions of China Electrotechnical Society, 1-11.
[9] 张玉旺,都义彬,张岩,等.无线电能传输系统参数辨识技术研究综述[J/OL]. 电源学报, 1-12 [2025-12-07].
Zhang Yuwang, Du Yibin, Zhang Yan, et al.A Review of the Research on Parameter Identification Technology of Wireless Power Transmission System [J/OL]. Journal of Power Supply, 1-12[2025-12-07].
[10] Luo Bo, Wu Huan, Wang Mengyao, et al.Front-End Parameter Identification Method Based on Adam-W Optimization Algorithm for Underwater Wireless Power Transfer System[J], IEEE Transactions on Power Electronics, 2025, 40(4): 6307-6318.
[11] Su Yugang, Zhang Hongyan, Wang Zhihui,et al.Steady-state load identification method of inductive power transfer system based on switching capacitors[J].IEEE Transactions on Power Electronics,2015,30(11): 6349-6355.
[12] Yin Jian, Lin Deyan, Thomas ParisiniLin, et al. Front-End Monitoring of the Mutual Inductance and Load Resistance in a Series-Series Compensated Wireless Power Transfer System[J], IEEE Transactions on Power Electronics,2016, 31(10): 7339-7352.
[13] Zeng Junming, Yang Yun, Li Kerui, et al.An ultrafast estimation method for coupling coefficient and receiver resonant frequency in universal wireless power transfer systems[J], IEEE Transactions on Power Electronics, 2024, 39(4): 4870-4883.
[14] Yang Yun, Tan Siew-Chong, Shu Yuen Ron Hui. Front-End Parameter Monitoring Method Based on Two-Layer Adaptive Differential Evolution for SS-Compensated Wireless Power Transfer Systems[J], IEEE Transactions on Industrial Informatics, 2019, 15(11): 6101-6113
[15] 刘廷章,张子恒.基于循环迭代粒子群算法对电动汽车无线充电系统的联合参数在线辨识[J].电工技术学报,2022,37(18):4548-4564.
Liu Tingzhang, Zhang Ziheng.Online Identification of Joint Parameters of Electric Vehicle Wireless Charging System Based on Loop-Iteration Particle Swarm Optimization[J]. Transactions of China Electrotechnical Society, 2022, 37(18): 4548-4564.
[16] Zhang Ben, Yang Fengshuo, Zhou Jiayu, A High-Compatibility Wireless Power Transfer System for Autonomous Underwater Vehicles Based on Reconfigurable Magnetic Couplers[J], IEEE Transactions on Industry Applications, 2025,61(2): 3457-3467.
[17] 王得安,张剑韬,朱春波,等.海洋环境对水下无线电能传输系统的影响机理研究进展[J].电工技术学报,2025,40(03):653-675.
Wang De'an, Zhang Jiantao, Zhu Chunbo, et al. Review of Progress in the Study of Marine Environment Effects on Underwater Wireless Power Transfer Systems[J]. Transactions of China Electrotechnical Society, 2025, 40(03): 653-675.
[18] 刘旭,钟敬,荣灿灿,等.考虑水下跨接电容效应的水下磁耦合谐振式无线电能传输系统多参数辨识方法[J].电工技术学报,2026,41(02):374-388.
Liu Xu, Zhong Jing, Rong Cancan, et al.Multiple Parameters Identification Method with Considering the Span Capacitor Effect in the Underwater Magnetic Coupled Resonant Wireless Power Transfer System[J]. Transactions of China Electrotechnical Society, 2026, 41(02): 374-388.
[19] 李陶,王懿杰,李阳,等.面向水下无线电能传输系统涡流损耗抑制与效率优化的接收端控制方法研究[J].中国电机工程学报,2026,46(04):1572-1581.
Li Tao, Wang Yijie, Li Yang, et al.Research on Receiver-Side Control Methods for Eddy Current Loss Suppression and Efficiency Optimization in Underwater Wireless Power Transfer Systems[J]. Proceedings of the CSEE, 2026, 46(04): 1572-1581.
[20] Wang De'an, Zhang Jiantao, Cui Shumei, et al. A Comprehensive Review and New Perspectives on the State-of-the-Arts of Underwater Wireless Power Transfer[J]. Renewable and Sustainable Energy Reviews, 2024, 189(Part A): 113910.
[21] Yan Zhengchao, Zhang Yiming, Kan Tianze Frequency Optimization of a Loosely Coupled Underwater Wireless Power Transfer System Considering Eddy Current Loss[J], IEEE Transactions on Industrial Electronics, 2019,66(5):3468-3476.
[22] Liu Xu, Lu Damin, Zhijuan Liao et al. Critical and Parasitic Parameters Identification and Frequency Regulation Strategies for UWPT Systems[J], IEEE Transactions on Power Electronics, 2024,39(8): 10558-10567.
[23] Chen Zhixin, Yang Qingxin, Zhang Xian, et al. Hybrid Mechanism- and Data-Driven Joint Parameter Identification for Underwater Wireless Power Transfer Systems[J], IEEE Transactions on Power Electronics, 2025.3616222.
[24] Chen Tianqi and Carlos Guestrin. XGBoost: A Scalable Tree Boosting System. Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (2016).
[25] 白龙雷,孙良顺,马智勇,等.基于原边监测的水下WPT系统参数识别技术研究[J].中国电机工程学报,2025,45(06):2358-2367.
Bai Longlei, Sun Liangshun, Ma Zhiyong et al. Research on Parameter Identification Technology of Underwater WPT System Based on Transmission Side Monitoring[J]. Proceedings of the CSEE, 2025,45(06):2358-2367.