运用经验模态分解（EMD）将某大跨度膜结构测点非平稳风压分解为一系列相对平稳的固有模态函数和一个剩余分量.为消除实测风压中噪声对固有模态函数的影响，使用小波变换对每个固有模态函数进行去噪，将去噪后的固有模态函数及剩余分量作为样本输入.分别将径向基核函数、Hermite核函数及Hermite组合核与最小二乘支持向量机结合(LSSVM)，运用粒子群算法(PSO)对3种算法的正则化参数及核参数进行智能寻优，建立基于径向基核函数、Hermite核函数及Hermite组合核的PSO-LSSVM风压预测算法，并基于超高层建筑实测风压验证了组合模型的鲁棒性.单点预测结果表明，基于Hermite组合核的PSO-LSSVM的预测算法较其余两种算法具有更高预测精度及泛化能力；空间点预测结果进一步证明了该方法对于非平稳非高斯风压预测的有效性.
Empirical mode decomposition (EMD) is used to decompose the non-stationary wind pressure of a long-span membrane structure into a series of relatively stationary intrinsic mode functions and a residual component. In order to eliminate the effect of noise on the intrinsic mode function in actual wind pressure measurement, each intrinsic mode function is denoised by using wavelet transform. The residual components and the intrinsic mode functions after denoising are input as samples. Radial basis kernel, Hermite kernel and Hermite combination kernel are combined with least square support vector machine (LSSVM) respectively. Subsequently, optimizations for penalty parameters and kernel parameters are conducted using particle swarm optimization (PSO) and thus three algorithms based on PSO-LSSVM are proposed for wind pressure prediction. In addition, the robustness of the combined model is verified based on the measured wind pressure on the super high-rise building. The single point predicting indexes show that the prediction algorithm using Hermite combination kernel based on PSO-LSSVM has higher prediction accuracy and generalization ability compared to the other two algorithms. The results of spatial point prediction further prove the validity of this method for non-stationary non-Gaussian wind pressure prediction.
［1］YUAN X H, CHEN C, YUAN Y B, et al. Short-term wind power prediction based on LSSVM-GSA model［J］. Energy Conversion and Management, 2015, 101: 393-401.
［2］DE GIORGI M, CAMPILONGO S, FICARELLA A, et al. Comparison between wind power prediction models based on wavelet decomposition with least-squares support vector machine (LS-SVM) and artificial neural network (ANN)［J］. Energies, 2014, 7(8): 5251-5272.
［3］WANG H, WU T, TAO T Y, et al. Measurements and analysis of non-stationary wind characteristics at sutong bridge in typhoon damrey［J］. Journal of Wind Engineering and Industrial Aerodynamics, 2016, 151: 100-106.
［4］吴本刚, 傅继阳, 吴玖荣. 实测风场风速风向耦合的三维非平稳特征研究［J］. 建筑结构学报, 2016, 37(2): 106-113.
WU Bengang, FU Jiyang, WU Jiurong. Three dimensional non-stationary analysis on field measured wind data with coupling wind speed and wind direction［J］. Journal of Building Structures, 2016, 37(2): 106-113.
［5］孙旭峰, BITSUAMLAK G T, 胡超. 屋盖结构脉动风压非高斯特性分析的极限流线方法［J］. 振动与冲击, 2015, 34(8): 157-162.
SUN Xufeng, BITSUAMLAK G T, HU Chao. Li-miting streamline method for analysis of non-Gaussian property of roof structures’ fluctuating wind pressure［J］. Journal of Vibration and Shock, 2015, 34(8): 157-162.
［6］DYBAA J, ZIMROZ R. Rolling bearing diagnosing method based on empirical mode decomposition of machine vibration signal［J］. Applied Acoustics, 2014, 77: 195-203.
［7］KATICHA S W, FLINTSCH G, BRYCE J, et al. Wavelet denoising of TSD deflection slope measurements for improved pavement structural evaluation［J］. Computer-Aided Civil and Infrastructure Engineering, 2014, 29(6): 399-415.
［8］SRIVASTAVA M, GEORGIEVA E R, FREED J H. A new wavelet denoising method for experimental time-domain signals: pulsed dipolar electron spin re-sonance［J］. The Journal of Physical Chemistry A, 2017, 121(12): 2452-2465.
［9］SU J Q, WANG X, LIANG Y, et al. GA-based support vector machine model for the prediction of monthly reservoir storage［J］. Journal of Hydrologic Engineering, 2014, 19(7): 1430-1437.
［10］AHMADI M A, BAHADORI A. A LSSVM approach for determining well placement and conning phenomena in horizontal wells［J］. Fuel, 2015, 153: 276-283.
［11］HOOSHMAND MOGHADDAM V, HAMIDZADEH J. New hermite orthogonal polynomial kernel and combined kernels in support vector machine classifier［J］. Pattern Recognition, 2016, 60: 921-935.
［12］张志宏, 刘中华, 董石麟. 强/台风作用下大跨空间索桁体系现场风压风振实测研究［J］. 上海师范大学学报(自然科学版), 2013, 42(5): 546-550.
ZHANG Zhihong, LIU Zhonghua, DONG Shilin. Field measurement of wind pressure and wind-induced vibration of large-span spatial cable-truss system under strong wind or typhoon［J］. Journal of Shanghai Normal University (Natural Sciences), 2013, 42(5): 546-550.
［13］SUBASI A. Classification of EMG signals using PSO optimized SVM for diagnosis of neuromuscular disorders［J］. Computers in Biology and Medicine, 2013, 43(5): 576-586.
［14］SELAKOV A, CVIJETINOVI D, MILOVI L, et al. Hybrid PSO-SVM method for short-term load forecasting during periods with significant temperature variations in city of Burbank［J］. Applied Soft Computing, 2014, 16: 80-88.
