基于小波支持向量机的非高斯空间风压内外插预测

doi:10.16183/j.cnki.jsjtu.2018.11.014

摘要/Abstract

摘要： 支持向量机(SVM)的性能取决于核函数及核参数的选取.基于小波分析理论构造出满足Mercer平移不变核定理的Mexican Hat小波核函数(MW)，将MW和B样条核函数分别与最小二乘支持向量机(LSSVM)结合，形成MW-LSSVM和BS-LSSVM.运用粒子群(PSO)算法对MW-LSSVM和BS-LSSVM的正则化参数及核参数进行智能优化，建立了PSO-MW-LSSVM和PSO-BS-LSSVM的空间风压预测算法.实测风压预测结果表明，MW-LSSVM比BS-LSSVM和传统的径向基核函数RBF-LSSVM具有更好的非高斯风压预测性能及泛化能力，而且稳定性更强，具有较高的工程应用价值.

关键词: 小波核函数, 样条核函数, 最小二乘支持向量机, 风压预测, 粒子群算法

Abstract: The performance of support vector machine depends on the selection of kernel functions and kernel parameters. The Mexican Hat wavelet kernel function is constructed based on the wavelet analysis theory which could satisfy the Mercer conditions. Further, the Mexican Hat wavelet kernel function and the B-spline kernel are combined with the LSSVM respectively and accordingly MW-LSSVM and BS-LSSVM are proposed. Subsequently, optimizations for penalty parameters and kernel parameters are conducted using particle swarm optimization (PSO) and thus the PSO-MW-LSSVM and PSO-BS-LSSVM algorithms are proposed for spatial wind pressure prediction. The numerical analysis shows that the proposed method not only significantly outperforms the conventional RBF-LSSVM and BS-LSSVM in forecasting accuracy and generalization ability, but also has great engineering application prospects due to its stability.

Key words: wavelet kernel function, spline kernel function, least square support vector machine, wind pressure prediction, particle swarm

中图分类号:

TU 311

李春祥,殷潇. 基于小波支持向量机的非高斯空间风压内外插预测[J]. 上海交通大学学报（自然版）, 2018, 52(11): 1516-1523.

LI Chunxiang,YIN Xiao. Interpolation Prediction and Extrapolation Prediction of Non-Gaussian Spatial Wind Pressure Using LSSVM with Wavelet Kernel Function[J]. Journal of Shanghai Jiaotong University, 2018, 52(11): 1516-1523.

参考文献

［1］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. ［2］DING J, CHEN X Z. Assessment of methods for extreme value analysis of non-Gaussian wind effects with short-term time history samples［J］. Engineering Structures, 2014, 80: 75-88. ［3］QU L, ZHOU H. The multi-class SVM is applied in transformer fault diagnosis［C］//International Symposium on Distributed Computing and Applications for Business Engineering and Science. Paris: IEEE, 2016: 477-480. ［4］YAN X, CHOWDHURY N A. Mid-term electricity market clearing price forecasting: A hybrid LSSVM and ARMAX approach［J］. International Journal of Electrical Power & Energy Systems, 2013, 53(1): 20-26. ［5］SU J, WANG X, LIANG Y, et al. GA-based support vector machine model for the prediction of monthly reservoir storage［J］. Journal of Hydrologic Engineering, 2013, 19(7): 1430-1437. ［6］AHMADI M A, BAHADORI A. A LSSVM approach for determining well placement and conning phenomena in horizontal wells［J］. Fuel, 2015, 153: 276-283. ［7］SHANG F H, MIAO X J, WANG Z Y, et al. Automatic identifying algorithm of water-flooded zone based on B-SVM［C］//International Conference on Machine Learning and Cybernetics. Dalian: IEEE, 2006: 4035-4039. ［8］王春枝, 张会丽, 叶志伟. 基于混沌粒子群算法和小波SVM的P2P流量识别方法［J］. 计算机科学, 2015, 42(10): 117-121. WANG Chunzhi, ZHANG Huili, YE Zhiwei. Peer-to-peer traffic identification method based on chaos particle swarm algorithm and wavelet SVM［J］. Computer Science, 2015, 42(10): 117-121. ［9］迟恩楠, 李春祥. 基于优化组合核和Morlet小波核的LSSVM脉动风速预测方法［J］. 振动与冲击, 2016, 35 (18): 52-57. CHI Ennan, LI Chunxiang. Forecast of fluctuating wind velocity using LSSVM with optimized combination kernel and Morlet wavelet kernel［J］. Journal of Vibration and Shock, 2016, 35 (18): 52-57. ［10］LIU Z, CUI Y, LI W. A classification method for complex power quality disturbances using EEMD and rank wavelet SVM［J］. IEEE Transactions on Smart Grid, 2017, 6(4): 1678-1685. ［11］ZHOU J Y, JING S, GONG L. Fine tuning support vector machines for short-term wind speed forecasting［J］. Energy Conversion and Management, 2011, 52(4): 1990-1998. ［12］SUBASI A. Classification of EMG signals using PSO optimized SVM for diagnosis of neuromuscular dis-orders［J］. Computers in Biology & Medicine, 2013, 43(5): 576-586. ［13］SELAKOV A, MELLON S, BEKUT D. 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(3): 80-88. ［14］李锦华, 吴春鹏, 陈水生. 矩形结构非高斯风荷载特性研究［J］. 振动、测试与诊断, 2014, 34(5): 951-959. LI Jinhua, WU Chunpeng, CHEN Shuisheng. Cha-racteristics of non-gaussian wind pressures on rectangular structure［J］. Journal of Vibration, Measurement & Diagnosis, 2014, 34(5): 951-959.

[1]	符杨, 丁枳尹, 米阳. 计及储能调节的时滞互联电力系统频率控制[J]. 上海交通大学学报, 2022, 56(9): 1128-1138.
[2]	李春祥，裴杨从琪，殷潇. 基于Hermite组合核EMD-WT-LSSVM的非平稳非高斯风压预测[J]. 上海交通大学学报, 2019, 53(10): 1249-1258.
[3]	徐峰1，范春菊1，徐勋建2，李丽2，倪佳筠3. 基于变分模态分解和AMPSO-SVM耦合模型的滑坡位移预测[J]. 上海交通大学学报（自然版）, 2018, 52(10): 1388-1395.
[4]	张朝飞，罗建军，徐兵华，马卫华. 基于灰色理论的新陈代谢自适应多参数预测方法 [J]. 上海交通大学学报（自然版）, 2017, 51(8): 970-976.
[5]	胡新明,王德禹. 基于迭代均值组合近似模型和序贯优化与可靠性评估法的船舶结构优化设计[J]. 上海交通大学学报（自然版）, 2017, 51(2): 150-.
[6]	罗华毅，王景成，杨丽雯，李肖城. 基于时差系数的城市原水需水量预测应用[J]. 上海交通大学学报（自然版）, 2017, 51(10): 1260-1267.
[7]	马驰,赵亮,梅雪松,施虎,杨军. 基于粒子群算法与BP网络的机床主轴热误差建模[J]. 上海交通大学学报（自然版）, 2016, 50(05): 686-695.
[8]	杨军,梅雪松,赵亮,马驰,冯斌,施虎. 基于模糊聚类测点优化与向量机的坐标镗床热误差建模[J]. 上海交通大学学报（自然版）, 2014, 48(08): 1175-1182.
[9]	姜文英，林焰，陈明，于雁云. 基于粒子群和蚁群算法的船舶机舱规划方法[J]. 上海交通大学学报（自然版）, 2014, 48(04): 502-507.
[10]	杜继永1, 2，张凤鸣1，惠晓滨1，李永宾1. 改进型连续粒子群算法求解重叠联盟生成问题[J]. 上海交通大学学报（自然版）, 2013, 47(12): 1918-1923.
[11]	何小二,王德禹,夏利娟. 基于粒子群算法的多用途船结构优化[J]. 上海交通大学学报（自然版）, 2013, 47(06): 928-931.
[12]	张志英1, 杨克开1, 于瑾维2. 面向船体分段建造的二维不规则空间调度方法[J]. 上海交通大学学报（自然版）, 2012, 46(04): 651-656.
[13]	苏永生, 王永生, 段向阳. 实船喷水推进泵空化识别试验[J]. 上海交通大学学报（自然版）, 2012, 46(03): 404-409.
[14]	张广明, 袁宇浩, 龚松建. 基于改进最小二乘支持向量机方法的短期风速预测[J]. 上海交通大学学报（自然版）, 2011, 45(08): 1125-1129.
[15]	邓卫卫1, 杨慧中1, 2. 一种带监督的仿射传播聚类多模型建模方法[J]. 上海交通大学学报（自然版）, 2011, 45(08): 1172-1175.