基于小波和乘法混合核函数LSSVM的顺风向非高斯空间风压预测
发布时间:2019-05-06 18:59
【摘要】:提出了基于Marr小波核函数最小二乘支持向量机(Marr-LSSVM)的顺风向非高斯空间风压预测算法。通过传统高斯核函数(RBF)和多项式核函数(Poly)的乘法运算,提出了Poly*RBF-LSSVM(MK-LSSVM)的空间风压预测算法。运用粒子群优化(PSO)算法,对Marr-LSSVM、传统单核CSK-LSSVM和MK-LSSVM的惩罚参数、核函数参数、权重、尺度因子进行优化,建立基于智能优化的非高斯空间风压预测算法;以30 m和50 m处模拟顺风向风压时程作为输入样本,使用提出的预测算法对40 m处风压时程进行了预测。数值分析表明,Marr-LSSVM、MK-LSSVM比CSK-LSSVM具有明显高的非高斯风压预测性能。
[Abstract]:Based on the Marr wavelet kernel function least squares support vector machine (Marr-LSSVM), a forward wind pressure prediction algorithm for non-Gao Si space is proposed. Based on the multiplication of traditional Gao Si kernel function (RBF) and polynomial kernel function (Poly), a spatial wind pressure prediction algorithm for Poly*RBF-LSSVM (MK-LSSVM) is proposed. Particle swarm optimization (PSO) algorithm is used to optimize the penalty parameters, kernel function parameters, weights and scale factors of traditional single-core CSK-LSSVM and MK-LSSVM in Marr-LSSVM, and a non-Gaussian spatial wind pressure prediction algorithm based on intelligent optimization is established. Taking 30 m and 50 m as input samples, the time history of wind pressure at 40 m is predicted by using the proposed prediction algorithm. Numerical analysis shows that Marr-LSSVM,MK-LSSVM has significantly higher non-Gao Si wind pressure prediction performance than CSK-LSSVM.
【作者单位】: 上海大学土木工程系;同济大学建筑工程系;
【基金】:国家自然科学基金(51378304)
【分类号】:TP18
本文编号:2470410
[Abstract]:Based on the Marr wavelet kernel function least squares support vector machine (Marr-LSSVM), a forward wind pressure prediction algorithm for non-Gao Si space is proposed. Based on the multiplication of traditional Gao Si kernel function (RBF) and polynomial kernel function (Poly), a spatial wind pressure prediction algorithm for Poly*RBF-LSSVM (MK-LSSVM) is proposed. Particle swarm optimization (PSO) algorithm is used to optimize the penalty parameters, kernel function parameters, weights and scale factors of traditional single-core CSK-LSSVM and MK-LSSVM in Marr-LSSVM, and a non-Gaussian spatial wind pressure prediction algorithm based on intelligent optimization is established. Taking 30 m and 50 m as input samples, the time history of wind pressure at 40 m is predicted by using the proposed prediction algorithm. Numerical analysis shows that Marr-LSSVM,MK-LSSVM has significantly higher non-Gao Si wind pressure prediction performance than CSK-LSSVM.
【作者单位】: 上海大学土木工程系;同济大学建筑工程系;
【基金】:国家自然科学基金(51378304)
【分类号】:TP18
【相似文献】
相关期刊论文 前2条
1 王宏健;徐金龙;李娟;张爱华;;非平稳非高斯测量噪声条件下改进差分粒子滤波算法研究[J];兵工学报;2014年07期
2 ;[J];;年期
相关博士学位论文 前2条
1 王刚;基于最大非高斯估计的独立分量分析理论研究[D];国防科学技术大学;2005年
2 屈毅;非高斯随机分布系统控制与故障检测方法的研究[D];兰州理工大学;2013年
相关硕士学位论文 前2条
1 杨楠;基于改进核独立元分析的非高斯过程监测方法研究[D];东北大学;2014年
2 许大星;非线性系统的几种滤波算法研究[D];杭州电子科技大学;2014年
,本文编号:2470410
本文链接:https://www.wllwen.com/kejilunwen/zidonghuakongzhilunwen/2470410.html