基于改进回声状态网络的质子交换膜燃料电池剩余寿命预测

doi:10.11930/j.issn.1004-9649.202402054

摘要/Abstract

摘要：

针对质子交换膜燃料电池（PEMFC）的剩余有效寿命预测技术（RUL）在中长期预测效果不佳的问题，提出了一种基于改进灰狼优化算法（IGWO）和回声状态网络（ESN）的剩余寿命预测方法。首先选取电堆电压作为健康指标，使用卷积平滑滤波法对PEMFC数据集进行数据平滑和归一化处理，有效减少异常值对后续模型训练的干扰。然后利用IGWO的局部和全局寻优能力对ESN的储备池参数进行优化，构建出IGWO-ESN网络模型，并利用处理后数据集进行PEMFC剩余寿命预测模型的训练，最后与传统的ESN进行对比验证。结果表明，改进后的ESN模型预测均方根误差和平均绝对百分比误差分别为0.0342和0.9315%，预测精度相较于普通ESN模型明显提升，中长期RUL的预测准确度也更高。

关键词: 质子交换膜燃料电池, 回声状态网络, 灰狼优化算法, 剩余寿命预测

Abstract:

Aiming at the problem that the current residual effective life prediction (RUL) technique for proton exchange membrane fuel cells (PEMFCs) has poor prediction effect in the medium and long term, a residual life prediction method based on the Improved Gray Wolf Optimization algorithm (IGWO) and Echo State Network (ESN) is proposed, in which the voltage of the electric stack is firstly selected as a health indicator, and the PEMFC dataset is processed by using convolutional smoothing filtering method to carry out data Smoothing and normalization are used to effectively reduce the interference of outliers on the subsequent model training. Then the reserve pool parameters of the ESN are optimized using the local and global optimization search capability of IGWO, and the IGWO-ESN network model is constructed, and the processed dataset is used for the training of the remaining life prediction model of the PEMFC, and finally it is compared with the traditional ESN for verification. The results show that the improved ESN model predicts the root mean square error and average absolute percentage error of 0.0342 and 0.9315%, respectively, and the prediction accuracy is significantly improved compared with the ordinary ESN model, and the prediction accuracy of the medium- and long-term RUL is also higher.

Key words: proton exchange membrane fuel cell, echo state network, gray wolf optimization algorithm, remaining life prediction

袁铁江, 李荣盛, 康建东, 闫华光. 基于改进回声状态网络的质子交换膜燃料电池剩余寿命预测[J]. 中国电力, 2025, 58(5): 102-109.

YUAN Tiejiang, LI Rongsheng, KANG Jiandong, YAN Huaguang. Residual Life Prediction of Proton Exchange Membrane Fuel Cell Based on Improved ESN[J]. Electric Power, 2025, 58(5): 102-109.

导出引用管理器 EndNote|Ris|BibTeX

链接本文: https://www.electricpower.com.cn/CN/10.11930/j.issn.1004-9649.202402054

https://www.electricpower.com.cn/CN/Y2025/V58/I5/102

图/表 11

图 1 回声状态网络结构示意

Fig.1 Echo state network structure

图 2 IGWO-ESN算法流程

Fig.2 IGWO-ESN algorithm flow

表 1 PEMFC关键参数

Table 1 Table 1 PEMFC key parameters

物理意义		参数
U1单电池电压/V		V₁
U2单电池电压/V		V₂
U3单电池电压/V		V₃
U4单电池电压/V		V₄
U5单电池电压/V		V₅
电堆输出电压/V		U
燃料电池电流/A		I
氢气入口温度/℃		T_H2
空气入口湿度/%		H_r

图 3 部分传感器监测参数

Fig.3 Part of the sensor monitoring parameters

图 4 FC1电堆输出平滑电压

Fig.4 Output smooth voltage of FC1 stack

表 2 ESN和改进ESN模型具体参数

Table 2 Specific parameters of ESN and improved ESN model

模型	M	N	L	W_in、W	S_D	S_R
ESN	1	500	1	[–0.5, 0.5]	0.05	0.7
改进ESN	1	[100, 1000]	[–0.5, 0.5]	[–0.5, 0.5]	[0.01, 0.1]	[0.5, 1.5]

图 5 ESN谱半径与训练误差关系

Fig.5 Relationship between ESN spectral radius and training error

图 6 IGWO优化ESN迭代误差梯度

Fig.6 IGWO optimizes ESN iterative error gradient

图 7 中短期预测结果对比

Fig.7 Comparison of short-term prediction results

图 8 长期预测结果对比

Fig.8 Comparison of long-term prediction results

表 3 ESN和改进ESN模型评估

Table 3 ESN and improved ESN model evaluation

模型	时段	R_MSE	M_APE/%	S_core
ESN	1 h	0.0008	0.0212	1165
	10 h	0.0336	0.7731
	500 h	0.7563	21.0353
改进ESN	1 h	0.0006	0.0154	123
	10 h	0.0018	0.0495
	500 h	0.0342	0.9315

参考文献 29

1	舒印彪, 赵勇, 赵良, 等. “双碳” 目标下我国能源电力低碳转型路径[J]. 中国电机工程学报, 2023, 43 (5): 1663- 1672.
	SHU Yinbiao, ZHAO Yong, ZHAO Liang, et al. Study on low carbon energy transition path toward carbon peak and carbon neutrality[J]. Proceedings of the CSEE, 2023, 43 (5): 1663- 1672.
2	任大伟, 肖晋宇, 侯金鸣, 等. 双碳目标下我国新型电力系统的构建与演变研究[J]. 电网技术, 2022, 46 (10): 3831- 3839.
	REN Dawei, XIAO Jinyu, HOU Jinming, et al. Construction and evolution of China’s new power system under dual carbon goal[J]. Power System Technology, 2022, 46 (10): 3831- 3839.
3	邵志刚, 衣宝廉. 氢能与燃料电池发展现状及展望[J]. 中国科学院院刊, 2019, 34 (4): 469- 477.
	SHAO Zhigang, YI Baolian. Developing trend and present status of hydrogen energy and fuel cell development[J]. Bulletin of Chinese Academy of Sciences, 2019, 34 (4): 469- 477.
4	李慧敏, 涂淑平. 中国氢燃料电池技术发展现状、挑战及对策[J]. 现代化工, 2023, 43 (11): 5- 9.
	LI Huimin, TU Shuping. Challenges and countermeasures of key technologies for hydrogen fuel cells in China[J]. Modern Chemical Industry, 2023, 43 (11): 5- 9.
5	WANG L J, LI X Y, GUO P Y, et al. Bibliometric analysis of prognostics and health management (PHM) in hydrogen fuel cell engines[J]. International Journal of Hydrogen Energy, 2022, 47 (80): 34216- 34243. DOI
6	OU M Y, ZHANG R F, SHAO Z F, et al. A novel approach based on semi-empirical model for degradation prediction of fuel cells[J]. Journal of Power Sources, 2021, 488, 229435. DOI
7	ZHANG X F, YANG D J, LUO M H, et al. Load profile based empirical model for the lifetime prediction of an automotive PEM fuel cell[J]. International Journal of Hydrogen Energy, 2017, 42 (16): 11868- 11878. DOI
8	CHEN K, LAGHROUCHE S, DJERDIR A. Degradation model of proton exchange membrane fuel cell based on a novel hybrid method[J]. Applied Energy, 2019, 252, 113439. DOI
9	CHEN K, LAGHROUCHE S, DJERDIR A. Prognosis of fuel cell degradation under different applications using wavelet analysis and nonlinear autoregressive exogenous neural network[J]. Renewable Energy, 2021, 179, 802- 814. DOI
10	LIU H, CHEN J, HISSEL D, et al. Short-term prognostics of PEM fuel cells: a comparative and improvement study[J]. IEEE Transactions on Industrial Electronics, 2019, 66 (8): 6077- 6086. DOI
11	HUA Z G, ZHENG Z X, PAHON E, et al. A review on lifetime prediction of proton exchange membrane fuel cells system[J]. Journal of Power Sources, 2022, 529, 231256. DOI
12	HU X S, XU L, LIN X K, et al. Battery lifetime prognostics[J]. Joule, 2020, 4 (2): 310- 346. DOI
13	WANG F K, MAMO T, CHENG X B. Bi-directional long short-term memory recurrent neural network with attention for stack voltage degradation from proton exchange membrane fuel cells[J]. Journal of Power Sources, 2020, 461, 228170. DOI
14	ZHOU D M, AL-DURRA A, ZHANG K, et al. Online remaining useful lifetime prediction of proton exchange membrane fuel cells using a novel robust methodology[J]. Journal of Power Sources, 2018, 399, 314- 328. DOI
15	任圆圆, 许亮, 蔡远利. 质子交换膜燃料电池剩余使用寿命预测研究进展[J]. 电源技术, 2023, 47 (8): 984- 988. DOI
	REN Yuanyuan, XU Liang, CAI Yuanli. Progress on remaining useful life prediction of proton exchange membrane fuel cell[J]. Chinese Journal of Power Sources, 2023, 47 (8): 984- 988. DOI
16	LIU J W, LI Q, CHEN W R, et al. Remaining useful life prediction of PEMFC based on long short-term memory recurrent neural networks[J]. International Journal of Hydrogen Energy, 2019, 44 (11): 5470- 5480. DOI
17	PENG Y L, CHEN T, XIAO F, et al. Remaining useful lifetime prediction methods of proton exchange membrane fuel cell based on convolutional neural network-long short-term memory and convolutional neural network-bidirectional long short-term memory[J]. Fuel Cells, 2023, 23 (1): 75- 87. DOI
18	WANG F K, CHENG X B, HSIAO K C. Stacked long short-term memory model for proton exchange membrane fuel cell systems degradation[J]. Journal of Power Sources, 2020, 448, 227591. DOI
19	常家康, 吕宁, 詹跃东. 基于XGBoost-RFECV算法和LSTM神经网络的PEMFC剩余寿命预测[J]. 电子测量与仪器学报, 2022, 36 (1): 126- 133.
	CHANG Jiakang, LYU Ning, ZHAN Yuedong. Prediction of PEMFC remaining life based on XGBoost-RFECV algorithm and LSTM neural network[J]. Journal of Electronic Measurement and Instrumentation, 2022, 36 (1): 126- 133.
20	MORANDO S, PERA M C, STEINER N Y, et al. Fuel cells fault diagnosis under dynamic load profile using reservoir computing[C]//2016 IEEE Vehicle Power and Propulsion Conference (VPPC). Hangzhou, China. IEEE, 2016: 1–6.
21	LI Z L, JEMEI S, GOURIVEAU R, et al. Remaining useful life estimation for PEMFC in dynamic operating conditions[C]//2016 IEEE Vehicle Power and Propulsion Conference (VPPC). Hangzhou, China. IEEE, 2016: 1–6.
22	LI Z L, ZHENG Z X, OUTBIB R. Adaptive prognostic of fuel cells by implementing ensemble echo state networks in time-varying model space[J]. IEEE Transactions on Industrial Electronics, 2020, 67 (1): 379- 389. DOI
23	HUA Z G, ZHENG Z X, PÉRA M C, et al. Remaining useful life prediction of PEMFC systems based on the multi-input echo state network[J]. Applied Energy, 2020, 265, 114791. DOI
24	CHEN J, ZHOU D, LYU C, et al. A novel health indicator for PEMFC state of health estimation and remaining useful life prediction[J]. International Journal of Hydrogen Energy, 2017, 42 (31): 20230- 20238.
25	张春雁, 窦真兰, 王俊, 等. 电解水制氢-储氢-供氢在电力系统中的发展路线[J]. 发电技术, 2023, 44 (3): 305- 317.
	ZHANG Chunyan, DOU Zhenlan, WANG Jun, et al. Development route of hydrogen production by water electrolysis, hydrogen storage and hydrogen supply in power system[J]. Power Generation Technology, 2023, 44 (3): 305- 317.
26	倪筹帷, 陈杨, 张雪松, 等. 考虑安全性风险的电热氢系统优化配置方法[J]. 中国电力, 2024, 57 (9): 124- 135.
	NI Chouwei, CHEN Yang, ZHANG Xuesong, et al. Optimal configuration method for electric-thermo-hydrogen system considering safety risks[J]. Electric Power, 2024, 57 (9): 124- 135.
27	滕越, 赵骞, 袁铁江, 等. 绿电-氢能-多域应用耦合网络关键技术现状及展望[J]. 发电技术, 2023, 44 (3): 318- 330.
	TENG Yue, ZHAO Qian, YUAN Tiejiang, et al. Key technology status and outlook for green electricity-hydrogen energy-multi-domain applications coupled network[J]. Power Generation Technology, 2023, 44 (3): 318- 330.
28	冯兴, 杨威, 张安安, 等. 双向可逆的集中式电氢耦合系统容量优化配置[J]. 中国电力, 2024, 57 (8): 1- 11.
	FENG Xing, YANG Wei, ZHANG Anan, et al. Capacity optimization configuration of a bidirectional reversible centralized electrohydrogen coupling system[J]. Electric Power, 2024, 57 (8): 1- 11.
29	吴磊, 彭黎菊, 李爽, 等. 百千瓦级天然气制氢质子交换膜燃料电池热电联产系统稳态特性模拟分析[J]. 发电技术, 2023, 44 (3): 350- 360.
	WU Lei, PENG Liju, LI Shuang, et al. Simulation and analysis of steady state characteristics of hundred kilowatt proton exchange membrane fuel cell combined heat and power system based on hydrogen production from natural gas[J]. Power Generation Technology, 2023, 44 (3): 350- 360.

[1]	张娜, 任强, 刘广忱, 郭力萍, 李静宇. 基于VMD-GWO-ELMAN的光伏功率短期预测方法[J]. 中国电力, 2022, 55(5): 57-65.
[2]	张运厚, 李婉莹, 董福贵. 基于DE-GWO-SVR的中长期电力需求预测[J]. 中国电力, 2021, 54(9): 83-88.
[3]	左智科, 陈国彬, 刘超, 牛培峰. 基于CNGWO-LSSVM的汽轮机热耗率预测模型[J]. 中国电力, 2018, 51(8): 148-153.
[4]	赵洪山, 张健平, 高夺, 李浪. 风机齿轮箱轴承状态评估与剩余寿命预测[J]. 中国电力, 2017, 50(4): 141-145.