| 等式约束凸二次规划解析的新型神经网络方法* |
| |
| 引用本文： | 易称福,陈宇环,张小红.等式约束凸二次规划解析的新型神经网络方法*[J].计算机应用研究,2011,28(11):4023-4025. |
| |
| 作者姓名： | 易称福 陈宇环 张小红 |
| |
| 作者单位： | 1. 江西理工大学信息工程学院,江西赣州,341000
2. 赣南师范学院现代教育技术中心,江西赣州,341000 |
| |
| 基金项目： | 国家自然科学基金资助项目(11062002)；江西省自然科学基金资助项目(2010GZS0083) |
| |
| 摘 要： | 针对不同于传统基于梯度法的递归神经网络定义一种基于标量范数取值的非负能量函数，通过定义一种基于向量取值的不定无界的误差函数，构建了一种能实时求解具有线性等式约束的凸二次规划问题。基于Simulink仿真平台的计算机实验结果表明，该新型神经网络模型能够准确有效地求解此类二次规划问题。
|
| 关 键 词： | 递归神经网络 误差函数 梯度法 二次规划 |
Method of new neural network for solving convex quadratic programming with equality constraints |
| |
| YI Chen-fu,CHEN Yu-huan,ZHANG Xiao-hong.Method of new neural network for solving convex quadratic programming with equality constraints[J].Application Research of Computers,2011,28(11):4023-4025. |
| |
| Authors: | YI Chen-fu CHEN Yu-huan ZHANG Xiao-hong |
| |
| Affiliation: | (1. School of Information Engineering, Jiangxi University of Science & Technology, Ganzhou Jiangxi 341000, China; 2. Center for Educational Technology, Gannan Normal University, Ganzhou Jiangxi 341000, China) |
| |
| Abstract: | Differing from the conventional gradient-based recurrent neural networks, which associated with scalar-valued norm-based nonnegative energy function, by defining an indefinite error function based on the vector-valued, presented a new neural network for the online solution of the convex quadratic programming problem with equality constrains. Computer simulation results based on the Simulink show that the new neural networks can solve such quadratic programming with effectiveness and accuracy. |
| |
| Keywords: | recurrent neural networks error function gradient algorithm quadratic programming |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
| 点击此处可从《计算机应用研究》浏览原始摘要信息 |
|
点击此处可从《计算机应用研究》下载全文 |
