为提高光伏阵列故障诊断的准确性,提出一种结合改进蜜獾优化算法(IHBA)与双向长短期记忆网络(BiLSTM)的混合诊断模型。区别于面向功率预测的特征工程,该文聚焦于故障辨识,从光伏阵列的电流-电压与功率-电压特性曲线中,系统性提取涵盖基础、离散及分布统计的3类特征,形成10维度的综合特征向量。针对原始蜜獾算法易早熟收敛、搜索效率不足的缺陷,IHBA算法进行：采用Tent混沌映射改善种群初始分布、设计动态自适应控制因子以平衡搜索过程、引入小孔成像反向学习策略增强全局寻优能力3方面改进。基准函数测试表明,IHBA在收敛速度与求解精度上均优于对比算法。在此基础上,利用IHBA对BiLSTM网络的超参数进行自动寻优,可克服人工调参的盲目性,显著增强模型对高维非线性故障特征的建模能力与泛化性。最终,在包含正常、开路、短路、局部遮蔽及老化五类状态的仿真数据集上,IHBA-BiLSTM模型取得97.1014%的诊断准确率,其性能全面超越支持向量机、极限学习机、长短期记忆网络及其他智能优化算法结合的对比模型,证实该方法在光伏阵列多类故障诊断中兼具高精度与强鲁棒性。
Abstract
To improve the accuracy of fault diagnosis for photovoltaic （PV） arrays, this study proposes a hybrid fault diagnosis model combining the improved honey badger algorithm（IHBA） with the bidirectional long short-term memory （BiLSTM） network. Different from feature engineering oriented to power prediction, this paper focuses on fault identification, and systematically extracts three types of features covering basic, discrete and distributional statistics from the current-voltage and power-voltage characteristic curves of PV arrays, forming a 10-dimensional comprehensive feature vector.Aiming at the defects of the original Honey Badger Algorithm such as premature convergence and insufficient search efficiency, the IHBA is improved in three aspects: Tent chaos mapping is adopted to optimize the initial population distribution; a dynamic adaptive control factor is designed to balance the search process; and the pinhole imaging reverse learning strategy is introduced to enhance the global optimization ability. Benchmark function tests demonstrate that the IHBA outperforms the comparative algorithms in both convergence speed and solution accuracy. On this basis, the IHBA is utilized to automatically optimize the hyperparameters of the BiLSTM network, overcoming the blindness of manual parameter tuning and significantly enhancing the model’s modeling capability and generalization for high-dimensional nonlinear fault features. Finally, on the simulation dataset containing five operational states including normal, open-circuit, short-circuit, partial shading and aging, the IHBA-BiLSTM model achieves a fault diagnosis accuracy of 97.1014%. Its performance comprehensively surpasses that of the comparative models such as support vector machine （SVM）, extreme learning machine （ELM）, long short-term memory （LSTM） network and models integrated with other intelligent optimization algorithms, confirming that the proposed method possesses both high precision and strong robustness in the diagnosis of multiple types of faults in PV arrays.
关键词
故障诊断 /
特征提取 /
算法学习 /
光伏阵列 /
小孔成像策略 /
双向长短期记忆神经网络
Key words
fault detection /
feature extraction /
learning algorithms /
photovoltaic array /
pinhole imaging strategy /
bidirectional long short-term memory neural network
{{custom_sec.title}}
{{custom_sec.title}}
{{custom_sec.content}}
参考文献
[1] 贾凌云, 云斯宁, 赵泽妮, 等. 神经网络短期光伏发电预测的应用研究进展[J]. 太阳能学报, 2022, 43(12): 88-97.
JIA L Y, YUN S N, ZHAO Z N, et al.Recent progress of short-term forecasting of photovoltaic generation based on artificial neural networks[J]. Acta energiae solaris sinica, 2022, 43(12): 88-97.
[2] 王梦圆, 徐潇源, 严正. 适用小样本的并网光伏阵列故障诊断方法[J]. 电网技术, 2024, 48(2): 587-598.
WANG M Y, XU X Y, YAN Z.Simultaneous fault diagnosis of PV array operation considering small sample sizes[J]. Power system technology, 2024, 48(2): 587-598.
[3] 商立群, 闵鹏波, 张建涛. 基于改进PSO算法的光伏阵列MPPT研究[J]. 传感器与微系统, 2024, 43(8): 35-39.
SHANG L Q, MIN P B, ZHANG J T.Research on MPPT of photovoltaic array based on improved PSO algorithm[J]. Transducer and microsystem technologies, 2024, 43(8): 35-39.
[4] 吴琼, 解玉文, 郎庆凯, 等. 光伏组件智能化故障检测综述[J]. 上海电力大学学报, 2023, 39(3): 252-257.
WU Q, XIE Y W, LANG Q K, et al.Review of intelligent fault detection of photovoltaic modules[J]. Journal of Shanghai University of Electric Power, 2023, 39(3): 252-257.
[5] ZHANG J W, LIU Y J, LI Y L, et al.A reinforcement learning based approach for on-line adaptive parameter extraction of photovoltaic array models[J]. Energy conversion and management, 2020, 214: 112875.
[6] 蒋琳, 苏建徽, 李欣, 等. 基于可见光和红外热图像融合的光伏阵列热斑检测方法[J]. 太阳能学报, 2022, 43(1): 393-397.
JIANG L, SU J H, LI X, et al.Hot spot detection of photovoltaic array based on fusion of visible and infrared thermal images[J]. Acta energiae solaris sinica, 2022, 43(1): 393-397.
[7] 李光辉, 段晨东, 武珊. 基于半监督机器学习法的光伏阵列故障诊断[J]. 电网技术, 2020, 44(5): 1908-1913.
LI G H, DUAN C D, WU S.Fault diagnosis of PV array based on semi-supervised machine learning[J]. Power system technology, 2020, 44(5): 1908-1913.
[8] 任晓琳, 杨奕, 高龙, 等. 基于IBA-KELM模型的TCT光伏阵列故障诊断研究[J]. 电源学报, 2023, 21(5): 67-74.
REN X L, YANG Y, GAO L, et al.Research on fault diagnosis of TCT photovoltaic array based on IBA-KELM model[J]. Journal of power supply, 2023, 21(5): 67-74.
[9] 钱亮, 黄伟, 杨建卫. 基于HHO-ELM的光伏阵列故障诊断方法研究[J]. 电源技术, 2024, 48(2): 345-350.
QIAN L, HUANG W, YANG J W.Research on fault diagnosis method of photovoltaic array based on HHO-ELM[J]. Chinese journal of power sources, 2024, 48(2): 345-350.
[10] 赵靖英, 吴晶晶, 张雪辉, 等. 基于萤火虫扰动麻雀搜索算法-极限学习机的光伏阵列故障诊断方法研究[J]. 电网技术, 2023, 47(4): 1612-1625.
ZHAO J Y, WU J J, ZHANG X H, et al.Fault diagnosis of photovoltaic arrays based on sparrow search algorithm with firefly perturbation-extreme learning machine[J]. Power system technology, 2023, 47(4): 1612-1625.
[11] 李斌, 郭自强, 高鹏. 改进北方苍鹰算法在光伏阵列中应用研究[J]. 电子测量与仪器学报, 2023, 37(7): 131-139.
LI B, GUO Z Q, GAO P.Application of improved northern goshawk optimization algorithm in photovoltaic array[J]. Journal of electronic measurement and instrumentation, 2023, 37(7): 131-139.
[12] 顾崇寅, 徐潇源, 王梦圆, 等. 基于CatBoost算法的光伏阵列故障诊断方法[J]. 电力系统自动化, 2023, 47(2): 105-114.
GU C Y, XU X Y, WANG M Y, et al.CatBoost algorithm based fault diagnosis method for photovoltaic arrays[J]. Automation of electric power systems, 2023, 47(2): 105-114.
[13] 叶进, 卢泉, 王钰淞, 等. 基于级联随机森林的光伏故障诊断模型研究[J]. 太阳能学报, 2021, 42(3): 358-362.
YE J, LU Q, WANG Y S, et al.Research on PV fault diagnosis model based on cascaded random forest[J]. Acta energiae solaris sinica, 2021, 42(3): 358-362.
[14] ROLLA A, MOHAMED E A, SAMAH A, et al.Improving parameter estimation of fuel cell using honey badger optimization algorithm[J]. Frontiers in energy research, 2022, 10: 875332.
[15] 宋文乐, 张烨, 刘航旭, 等. 基于光伏组件老化特性解耦的渔光互补光伏短期出力预测方法[J]. 电力建设, 2024, 45(7): 25-33.
SONG W L, ZHANG Y, LIU H X, et al.Short-term output prediction method for complementary fishing and solar power decoupling the aging characteristics of photovoltaic modules[J]. Electric power construction, 2024, 45(7): 25-33.
[16] 秦宇, 许野, 王鑫鹏, 等. 基于改进FCM-LSTM的光伏出力短期预测研究[J]. 太阳能学报, 2024, 45(8): 304-313.
QIN Y, XU Y, WANG X P, et al.Study on short-term photovoltaic output prediction based on improved FCM-LSTM[J]. Acta energiae solaris sinica, 2024, 45(8): 304-313.
[17] PAN M Z, FU C C, CAO X X, et al.An energy management strategy for fuel cell hybrid electric vehicle based on HHO-BiLSTM-TCN-self attention speed prediction[J]. Energy, 2024, 307: 132734.
[18] DING D, LI J H, WANG H J, et al.ApplianceFilter: targeted electrical appliance disaggregation with prior knowledge fusion[J]. Applied energy, 2024, 365: 123157.
[19] 赵冬梅, 谢家康, 杜泽航, 等. 基于统计信息系数和Wasserstein生成对抗网络的风火系统暂态特征选择与两阶段稳定评估[J]. 电力自动化设备, 2023, 43(4): 106-113.
ZHAO D M, XIE J K, DU Z H, et al.Transient feature selection and two-stage stability assessment of wind-fire system based on uniform information coefficient and Wasserstein-generative adversarial network[J]. Electric power automation equipment, 2023, 43(4): 106-113.
[20] DAI H D, WANG J X, HUANG Y Y, et al.Lightweight state-of-health estimation of lithium-ion batteries based on statistical feature optimization[J]. Renewable energy, 2024, 222: 119907.
基金
国家重点研发计划（2022YFB2703500）; 云南省基础研究计划（202301AS070055）
PDF(1214 KB)
