期刊
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
卷 32, 期 1, 页码 36-48出版社
IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/TNNLS.2020.2973760
关键词
Optimization; Neurodynamics; Biological neural networks; Recurrent neural networks; Collaboration; Linear programming; Eigenvalues and eigenfunctions; Almost-sure convergence; mixed-integer optimization; neural networks
类别
资金
- Research Grants Council of the Hong Kong Special Administrative Region of China [11208517, 11202318]
- National Natural Science Foundation of China [61673330]
This article introduces a two-timescale duplex neurodynamic approach to mixed-integer optimization, utilizing two recurrent neural networks operating concurrently at two different timescales and employing particle swarm optimization for iterative updating of initial neuronal states. Despite its minimal system complexity, the method is almost surely convergent to optimal solutions and its superior performance is demonstrated through solving benchmark problems.
This article presents a two-timescale duplex neurodynamic approach to mixed-integer optimization, based on a biconvex optimization problem reformulation with additional bilinear equality or inequality constraints. The proposed approach employs two recurrent neural networks operating concurrently at two timescales. In addition, particle swarm optimization is used to update the initial neuronal states iteratively to escape from local minima toward better initial states. In spite of its minimal system complexity, the approach is proven to be almost surely convergent to optimal solutions. Its superior performance is substantiated via solving five benchmark problems.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据