| 不变凸非线性规划问题的神经网络模型 |
| |
| 引用本文： | 谢维,李国成. 不变凸非线性规划问题的神经网络模型[J]. 北京机械工业学院学报, 2012, 0(5): 58-63 |
| |
| 作者姓名： | 谢维 李国成 |
| |
| 作者单位： | 北京信息科技大学理学院,北京100192 |
| |
| 基金项目： | 国家自然科学基金（11101107） |
| |
| 摘 要： | 给出了解决不变凸非线性规划问题的神经网络模型。对于无约束问题,基于梯度下降法给出了递归神经网络模型,讨论了该模型的稳定性和优化性能;对于约束优化问题,基于逐次逼近的思想建立了一个反馈神经网络模型,并证明了该神经网络的收敛性。最后给出仿真实例验证神经网络的稳定性和优化能力。
|
| 关 键 词： | 不变凸非线性规划 递归神经网络 反馈神经网络 能量函数 |
Neural networks for solving the invex nonlinear programming problem |
| |
| XIE Wei,LI Guo-cheng. Neural networks for solving the invex nonlinear programming problem[J]. Journal of Beijing Institute of Machinery, 2012, 0(5): 58-63 |
| |
| Authors: | XIE Wei LI Guo-cheng |
| |
| Affiliation: | (School of Applied Sciences,Beijing Information Science and Technology University,Beijing 100192,China) |
| |
| Abstract: | This article discusses a neural network model to solve the invex nonlinear programming problem.Firstly,for the unconstrained problem,the recurrent neural network is given based on the gradient descent method,and the stability of the model and optimize performance are analyzed.Secondly,for the constrained optimization problem,establishing a feedback neural network model based on the idea of successive approximation and proving the convergence of the model.Finally,simulation examples are given to verify the stability and optimization capabilities of neural networks. |
| |
| Keywords: | invex nonlinear programming recurrent neural network feedback neural network energy function |
| 本文献已被 CNKI 维普 等数据库收录！ |
|
