| 一种基于径向基函数的近似模型构造方法 |
| |
| 引用本文： | 刘小民,张文斌.一种基于径向基函数的近似模型构造方法[J].燕山大学学报,2010,34(5). |
| |
| 作者姓名： | 刘小民 张文斌 |
| |
| 作者单位： | 西安交通大学能源与动力工程学院;中国船舶重工集团第704研究所; |
| |
| 基金项目： | 陕西省自然科学基金资助项目 |
| |
| 摘 要： | 发展了一种基于径向基函数的近似模型建立方法。首先采用试验设计方法对设计空间进行样本选择,采用数据分析方法对样本集进行聚类分类,得到径向基函数中的相关参数。对于近似模型权系数采用能够适应大规模并行计算并且能够在全局解空间的多个区域内寻优的遗传算法进行求解,而不是采用最小二乘法通过线性方程的求解获得。最后通过对一维非线性和多维高阶非线性函数的逼近测试,以及对近似模型的拟合函数与原函数之间的误差分析,表明本文提出的基于径向基函数的近似模型方法具有良好的拟合与预测精度。
|
| 关 键 词： | 径向基函数 近似模型 遗传算法 优化设计 |
An established scheme of surrogate model based on radial basis function |
| |
| LIU Xiao-min,ZHANG Wen-bin.An established scheme of surrogate model based on radial basis function[J].Journal of Yanshan University,2010,34(5). |
| |
| Authors: | LIU Xiao-min ZHANG Wen-bin |
| |
| Institution: | LIU Xiao-min 1,ZHANG Wen-bin 2 (1. School of Energy and Power Engineering,Xi'an Jiaotong University,Xi'an,Shaanxi 710049,China,2. Shanghai Marine Equipment Research Institute,China Shipbuilding Industry Corporation,Shanghai 200031,China) |
| |
| Abstract: | In this paper a surrogate model based on radial basis function (RBF) is presented. Firstly, design of expertiment (DOE) technology is adopted to select samples in design space. The sample set is classified by cluster analysis method and the parameters in the RBF are obtained. Then, the weight coefficients in the surrogate model are computed by genetic algorithm (GA) instead of least square method. Finally, two high-order nonlinearity function approximation problems are selected to test the performace of the... |
| |
| Keywords: | radial basis function surrogate model genetic algorithm optimal design |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
|
