Two-Timescale Multilayer Recurrent Neural Networks for Nonlinear Programming

Two-Timescale Multilayer Recurrent Neural Networks for Nonlinear Programming. Wang, J. & Wang, J. IEEE Transactions on Neural Networks and Learning Systems, 33(1):37–47, January, 2022.

Paper doi abstract bibtex

This article presents a neurodynamic approach to nonlinear programming. Motivated by the idea of sequential quadratic programming, a class of two-timescale multilayer recurrent neural networks is presented with neuronal dynamics in their output layer operating at a bigger timescale than in their hidden layers. In the two-timescale multilayer recurrent neural networks, the transient states in the hidden layer(s) undergo faster dynamics than those in the output layer. Sufﬁcient conditions are derived on the convergence of the two-timescale multilayer recurrent neural networks to local optima of nonlinear programming problems. Simulation results of collaborative neurodynamic optimization based on the two-timescale neurodynamic approach on global optimization problems with nonconvex objective functions or constraints are discussed to substantiate the efﬁcacy of the two-timescale neurodynamic approach.

@article{wang_two-timescale_2022,
	title = {Two-{Timescale} {Multilayer} {Recurrent} {Neural} {Networks} for {Nonlinear} {Programming}},
	volume = {33},
	issn = {2162-237X, 2162-2388},
	url = {https://ieeexplore.ieee.org/document/9240954/},
	doi = {10.1109/tnnls.2020.3027471},
	abstract = {This article presents a neurodynamic approach to nonlinear programming. Motivated by the idea of sequential quadratic programming, a class of two-timescale multilayer recurrent neural networks is presented with neuronal dynamics in their output layer operating at a bigger timescale than in their hidden layers. In the two-timescale multilayer recurrent neural networks, the transient states in the hidden layer(s) undergo faster dynamics than those in the output layer. Sufﬁcient conditions are derived on the convergence of the two-timescale multilayer recurrent neural networks to local optima of nonlinear programming problems. Simulation results of collaborative neurodynamic optimization based on the two-timescale neurodynamic approach on global optimization problems with nonconvex objective functions or constraints are discussed to substantiate the efﬁcacy of the two-timescale neurodynamic approach.},
	language = {en},
	number = {1},
	urldate = {2022-02-01},
	journal = {IEEE Transactions on Neural Networks and Learning Systems},
	author = {Wang, Jiasen and Wang, Jun},
	month = jan,
	year = {2022},
	keywords = {/unread},
	pages = {37--47},
}

Downloads: 0

{"_id":"TkiBBd7TxFqhtFzJM","bibbaseid":"wang-wang-twotimescalemultilayerrecurrentneuralnetworksfornonlinearprogramming-2022","author_short":["Wang, J.","Wang, J."],"bibdata":{"bibtype":"article","type":"article","title":"Two-Timescale Multilayer Recurrent Neural Networks for Nonlinear Programming","volume":"33","issn":"2162-237X, 2162-2388","url":"https://ieeexplore.ieee.org/document/9240954/","doi":"10.1109/tnnls.2020.3027471","abstract":"This article presents a neurodynamic approach to nonlinear programming. Motivated by the idea of sequential quadratic programming, a class of two-timescale multilayer recurrent neural networks is presented with neuronal dynamics in their output layer operating at a bigger timescale than in their hidden layers. In the two-timescale multilayer recurrent neural networks, the transient states in the hidden layer(s) undergo faster dynamics than those in the output layer. Sufﬁcient conditions are derived on the convergence of the two-timescale multilayer recurrent neural networks to local optima of nonlinear programming problems. Simulation results of collaborative neurodynamic optimization based on the two-timescale neurodynamic approach on global optimization problems with nonconvex objective functions or constraints are discussed to substantiate the efﬁcacy of the two-timescale neurodynamic approach.","language":"en","number":"1","urldate":"2022-02-01","journal":"IEEE Transactions on Neural Networks and Learning Systems","author":[{"propositions":[],"lastnames":["Wang"],"firstnames":["Jiasen"],"suffixes":[]},{"propositions":[],"lastnames":["Wang"],"firstnames":["Jun"],"suffixes":[]}],"month":"January","year":"2022","keywords":"/unread","pages":"37–47","bibtex":"@article{wang_two-timescale_2022,\n\ttitle = {Two-{Timescale} {Multilayer} {Recurrent} {Neural} {Networks} for {Nonlinear} {Programming}},\n\tvolume = {33},\n\tissn = {2162-237X, 2162-2388},\n\turl = {https://ieeexplore.ieee.org/document/9240954/},\n\tdoi = {10.1109/tnnls.2020.3027471},\n\tabstract = {This article presents a neurodynamic approach to nonlinear programming. Motivated by the idea of sequential quadratic programming, a class of two-timescale multilayer recurrent neural networks is presented with neuronal dynamics in their output layer operating at a bigger timescale than in their hidden layers. In the two-timescale multilayer recurrent neural networks, the transient states in the hidden layer(s) undergo faster dynamics than those in the output layer. Sufﬁcient conditions are derived on the convergence of the two-timescale multilayer recurrent neural networks to local optima of nonlinear programming problems. Simulation results of collaborative neurodynamic optimization based on the two-timescale neurodynamic approach on global optimization problems with nonconvex objective functions or constraints are discussed to substantiate the efﬁcacy of the two-timescale neurodynamic approach.},\n\tlanguage = {en},\n\tnumber = {1},\n\turldate = {2022-02-01},\n\tjournal = {IEEE Transactions on Neural Networks and Learning Systems},\n\tauthor = {Wang, Jiasen and Wang, Jun},\n\tmonth = jan,\n\tyear = {2022},\n\tkeywords = {/unread},\n\tpages = {37--47},\n}\n\n","author_short":["Wang, J.","Wang, J."],"key":"wang_two-timescale_2022","id":"wang_two-timescale_2022","bibbaseid":"wang-wang-twotimescalemultilayerrecurrentneuralnetworksfornonlinearprogramming-2022","role":"author","urls":{"Paper":"https://ieeexplore.ieee.org/document/9240954/"},"keyword":["/unread"],"metadata":{"authorlinks":{}},"html":""},"bibtype":"article","biburl":"https://bibbase.org/zotero/victorjhu","dataSources":["CmHEoydhafhbkXXt5"],"keywords":["/unread"],"search_terms":["two","timescale","multilayer","recurrent","neural","networks","nonlinear","programming","wang","wang"],"title":"Two-Timescale Multilayer Recurrent Neural Networks for Nonlinear Programming","year":2022}