To solve the problem of intermittent missing sensor data in wind pressure measurements, a method for predicting the wind pressure time series on the surface of structures by combining the improved empirical mode decomposition algorithm (IEMD) with bidirectional long short-term memory (BiLSTM) networks is proposed. First of all, intrinsic mode functions (IMFs) are obtained through the adaptive decomposition of wind pressure time series using the IEMD method with soft sifting stopping criterion. The sample entropy is adopted to reconstruct IMFs and obtain subsequences. Then, BiLSTM networks are established and trained with these subsequences, while the prediction is also conducted, in which the Bayesian algorithm is employed to optimize the hyperparameters of neural networks. Finally, a case study of wind load prediction based on the wind tunnel test data of low-rise buildings is conducted to validate the proposed model. The results show that, compared with the traditional prediction models (e.g. multi-layer perceptron, BiLSTM), the prediction model based on IEMD and BiLSTM presents higher accuracy and calculation efficiency, which is capable of predicting Gaussian and non-Gaussian wind pressure signals.
由于BiLSTM结构复杂,通过交叉验证方法确定各深度学习模型的超参数将耗费大量的时间成本,即便采用经验模态分解与样本熵相组合的时序预处理方法,训练时长仍会以子序列的总个数为单位成倍地增加,由此对复合模型的预测效率产生较大影响.因此本文引入贝叶斯优化算法率定深度学习模型的各项超参数,使得训练效率和预测精度得到大幅度提升,各测点总训练用时均小于3 h.
YANGQ S, CHENF X, ZHAOL,et al .Effects of upstream turbulence intensity on aerodynamic loads of low-rise buildings in atmospheric boundary layer flow using large eddy simulation[J].Engineering Mechanics,2021,38(12):25-38.(in Chinese)
LIC X, ZHANGH Y .Hybridizing multivariate empirical mode decomposition and extreme learning machine to predict non-stationary processes[J]. Journal of Shanghai Jiao Tong University,2020, 54(4): 376-386.(in Chinese)
[5]
LIL C, ZHOUH J, LIUH L, et al. A hybrid method coupling empirical mode decomposition and a long short-term memory network to predict missing measured signal data of SHM systems[J]. Structural Health Monitoring,2021,20(4):1778-1793.
[6]
CHENZ S, ZHANGL K, HUAJ M,et al .A framework of data-driven wind pressure predictions on bluff bodies using a hybrid deep learning approach[J]. Measurement and Control, 2023, 56(1/2): 237-256.
ZHENGX F, ZHONGW, LIC X. Multivariate nonstationary fluctuating wind velocity prediction using kernel-based extreme learning machine[J]. Journal of Vibration and Shock, 2017, 36(18): 223-230.(in Chinese)
[9]
CHENY, KOPPG A, SURRYD. Interpolation of wind-induced pressure time series with an artificial neural network[J]. Journal of Wind Engineering and Industrial Aerodynamics,2002,90(6):589-615.
[10]
HUANGD M, HES Q, HEX H,et al. Prediction of wind loads on high-rise building using a BP neural network combined with POD[J]. Journal of Wind Engineering and Industrial Aerodynamics, 2017,170:1-17.
[11]
CHENF B, WANGX L, LIX,et al. Prediction of wind pressures on tall buildings using wavelet neural network[J]. Journal of Building Engineering, 2022, 46: 103674.
[12]
GAVALDAX, FERRER-GENERJ, KOPPG A,et al .Interpolation of pressure coefficients for low-rise buildings of different plan dimensions and roof slopes using artificial neural networks[J]. Journal of Wind Engineering and Industrial Aerodynamics,2011,99(5):658-664.
[13]
BRE F, GIMENEZJ M, FACHINOTTIV D .Prediction of wind pressure coefficients on building surfaces using artificial neural networks[J]. Energy and Buildings,2018,158:1429-1441.
[14]
MALLICKM, MOHANTAA, KUMARA,et al .Prediction of wind-induced mean pressure coefficients using GMDH neural network[J]. Journal of Aerospace Engineering,2020,33(1):04019104.
CHIE N, LIC X, ZHENGX F .Forecasting non-Gaussian spatial along-wind pressure using wavelet kernel and multiplicative mixed kernel functions based LSSVM[J]. Journal of Vibration and Shock,2017,36(9):116-121.(in Chinese)
[19]
HUANGJ X, LIQ S, HANX L. Recovery of missing field measured wind pressures on a supertall building based on correlation analysis and machine learning[J]. Journal of Wind Engineering and Industrial Aerodynamics,2022,231:105237.
CHENF B, TANGB F, CAIQ R,et al .Wind load characteristics and wind pressure prediction of long-span flat roof[J].Journal of Vibration and Shock, 2021, 40(3): 226-232.(in Chinese)
[22]
LIT, WUT, LIUZ. Nonlinear unsteady bridge aerodynamics:reduced-order modeling based on deep LSTM networks[J].Journal of Wind Engineering and Industrial Aerodynamics,2020,198:104116.
[23]
LIMJ Y, KIMS, KIMH K,et al .Long short-term memory (LSTM)-based wind speed prediction during a typhoon for bridge traffic control[J]. Journal of Wind Engineering and Industrial Aerodynamics,2022,220:104788.
WUP, SUNJ W, FENGW B. Reduced order model based on autoencoder and long short-term memory network[J]. Acta Aerodynamica Sinica, 2021, 39(1): 73-81.(in Chinese)
[26]
LIJ H, ZHUD S, LIC X. Comparative analysis of BPNN, SVR,LSTM,Random Forest,and LSTM-SVR for conditional simulation of non-Gaussian measured fluctuating wind pressures[J]. Mechanical Systems and Signal Processing,2022, 178: 109285.
LIJ L, WANGY M, CHANGJ X, et al. Ultra-short-term predictions of wind power based on parallel machine learning[J].Journal of Hydroelectric Engineering, 2023, 42(1): 40-51.(in Chinese)
[29]
LIJ L, SONGZ H, WANGX F, et al. A novel offshore wind farm typhoon wind speed prediction model based on PSO-Bi-LSTM improved by VMD[J]. Energy,2022, 251: 123848.
[30]
LÜM Z, LIJ, NIUX S,et al. Novel deterministic and probabilistic combined system based on deep learning and self-improved optimization algorithm for wind speed forecasting[J]. Sustainable Energy Technologies and Assessments, 2022, 52:102186.
[31]
JASEENAK U, KOVOORB C .Decomposition-based hybrid wind speed forecasting model using deep bidirectional LSTM networks[J]. Energy Conversion and Management, 2021, 234:113944.
CUIJ X, YANGB. Survey on Bayesian optimization methodology and applications[J]. Journal of Software, 2018, 29(10): 3068-3090.(in Chinese)
[34]
HUANGN E, SHENZ, LONGS R,et al .The empirical mode decomposition and the Hilbert spectrum for nonlinear and non-stationary time series analysis[J]. Proceedings of the Royal Society of London Series A, 1998, 454(1971): 903-998.
[35]
LIUZ L, PENGD D, ZUOM J,et al. Improved Hilbert-Huang transform with soft sifting stopping criterion and its application to fault diagnosis of wheelset bearings[J]. ISA Transactions,2022,125: 426-444.
[36]
TANH C, XIES C, LIUR D,et al .Fatigue condition diagnosis of rolling bearing based on normalized balanced multiscale sample entropy[J]. International Journal of Fatigue, 2023, 172:107642.
[37]
HOCHREITERS, SCHMIDHUBERJ. Long short-term memory[J]. Neural Computation,1997,9(8): 1735-1780.