Optimal artificial neural network architecture design for modeling an industrial ethylene oxide plant

•A mixed integer nonlinear programming-based ANN formulation is proposed.•The method detects the optimum input variables and the ideal number of neurons.•The method synthesizes the optimal information flow through the feedforward ANN.•The modified training problem involves the number of connections...

Celý popis

Uložené v:
Podrobná bibliografia
Vydané v:Computers & chemical engineering Ročník 163; s. 107850
Hlavní autori: Sildir, Hasan, Sarrafi, Sahin, Aydin, Erdal
Médium: Journal Article
Jazyk:English
Vydavateľské údaje: Elsevier Ltd 01.07.2022
Predmet:
Artificial neural networks
Machine learning
Mixed integer nonlinear programming
Process modeling
Superstructure optimization
Superstructure optimization
Mixed integer nonlinear programming
Process modeling
Machine learning
Artificial neural networks
ISSN:0098-1354
On-line prístup:Získať plný text
Tagy: Pridať tag
Žiadne tagy, Buďte prvý, kto otaguje tento záznam!
Popis
Shrnutí:•A mixed integer nonlinear programming-based ANN formulation is proposed.•The method detects the optimum input variables and the ideal number of neurons.•The method synthesizes the optimal information flow through the feedforward ANN.•The modified training problem involves the number of connections in the optimization problem explicitly. Optimum selection of input variables, number of hidden neurons and connections among the network elements deliver the best configuration of an ANN, usually resulting in reduced over-fitting and improved test performance. This study focuses on the development of a superstructure-oriented feedforward ANN design and training algorithm whose impacts are demonstrated on an industrial Ethylene Oxide (EO) plant for the prediction of product related variables. Proposed method brings about a mixed integer nonlinear programming problem (MINLP) to be solved, which takes the existence of inputs, neurons, and connections among the network elements into account by binary variables in addition to continuous weights of existing connections. Our investigations show that almost 85% of the ANN connections are removed compared to the fully connected ANN (FC-ANN) with 50% decrease in the number of inputs of the ANN. The modified ANN delivers a better prediction performance over FC-ANN, since FC-ANN suffers from over-fitting.
ISSN:0098-1354
DOI:10.1016/j.compchemeng.2022.107850