A unified framework for the mathematical modelling, predictive analysis, and optimization of reaction systems using computational fluid dynamics, deep neural network and genetic algorithm: A case of butadiene synthesis

Application of neural network in metal adsorption using biomaterials (BMs): a review

“…91,92 To prevent the over-parameterization and over-training of the ANN system, researchers have advocated the implementation of emerging activation functions (e.g., SeLU, ReLU). 201,202 Using simple models (e.g., AdaBoost) ensures the generalizability of output in the small dataset. 203 It is commonly recognized that a smaller error value is better for optimum network learning; however, it may be possible that the network training stops due to getting stuck in local minima.…”

“…, SeLU, ReLU). 201,202 Using simple models ( e.g. , AdaBoost) ensures the generalizability of output in the small dataset.…”

Review of Machine Learning for Hydrodynamics, Transport, and Reactions in Multiphase Flows and Reactors

“…More recently, there is a growing tendency to use deep learning methods combined with optimization algorithms for multiphase reactor performance optimization. Gbadago et al proposed a novel integrated framework for flow and transport modeling, data analysis, and optimization of reactor systems utilizing CFD, DNN, and GA (Figure ). The optimization results outperformed those of other traditional methods while the optimization speed was remarkably boosted.…”

“…A unified CFD-DNN-GA framework for multitubular reactor design and optimization. Adapted with permission from ref . Copyright 2021 Elsevier.…”

“…After the model is well trained, it can be used as an approximator to perform a rapid sensitivity analysis of reaction conditions and find the optimal conditions to maximize the yield of the main products and minimize the conversion of side products. Moreover, the trained ML model can also be used as an objective function in an optimization function, such as GA to optimize reaction conditions …”