BPNNHDMR非线性近似模型的方法及的应用.docVIP

BPNNHDMR非线性近似模型的方法及的应用 　　文章编号：1674??2974(2014)05??0032??07?? 　　收稿日期? 　　基金项目：国家高新技术研究发展计划（863计划）资助项目（2012AA111802）?? 　　作者简介：李伟平（1971-），男，湖南邵阳人，湖南大学副教授，博士?? 　　?k通讯联系人，E??mail: lwpzlbb@yeah.cn?お? 　　摘 要： 提出基于误差反向传播神经网络（Back Propagation Neural Network, BPNN)的高维模型表示（high dimensional model representation, HDMR)方法，即BPNN??HDMR方法.BPNN??HDMR方法的显著优势是将BP神经网络的非线性函数逼近能力与高维模型的层级结构理论相结合来构建近似模型，并能够揭示输入变量之间固有的线性或非线性关系及其相关性，将构造模型复杂度由指数级增长降阶为多项式级，有效地解决了高维建模问题.通过测试和对比验证了该算法的效率和建模能力，并将该算法应用于矿用自卸车安全驾驶室翻车保护装置（Roll??Over Protective Structure, ROPS)的优化设计.通过优化结果验证了所提方法的可行性和有效性.?? 　　关键词：近似模型;高维模型;误差反向传播神经网络;非线性;结构优化 　　?? 　　中图分类号：U463.4 文献标识码：A 　　 　　 　　BPNN??HDMR Nonlinear Metamodeling Technique?? 　　 and Its Application 　　?お? 　　LI Wei??ping???k, DOU Xian??dong, WANG Zhen??xing, LIU Chao?? 　　（State Key Laboratory of Advanced Designed and Manufacture for Vehicle Body,?? 　　 Hunan Univ, Changsha, Hunan 410082,China） 　　Abstract:This paper proposed a new high??dimension model representation (HDMR) based on back propagation neural network (BPNN), which is called BPNN??HDMR. The most remarkable advantage of this method lies in its ability to integrate the nonlinear function approximation capability of BP neural network and the hierarchy structure theory of high dimensional model to build an approximation model. Moreover, this method can reveal the inherent linearity or nonlinearity relationship as well as correlation with respect to input variables. The problem of modeling high dimension model is effectively tackled by reducing the computation cost from exponential growing to polynomial. Testing and comparative analysis confirm the efficiency and capability of BPNN??HDMR for high dimension nonlinear problems. Furthermore, the algorithm was applied to optimize the ROPS of Mining Dump Trucks Safety Cab. The optimized results verify the feasibility and effectiveness of the method proposed.?? 　　Key words: meta??model; high dimensional model representation（HDMR）; back propagatio