Metacontrol: A Python based application for self-optimizing control using metamodels

Application of Surrogate Models as an Alternative to Process Simulation for Implementation of the Self-Optimizing Control Procedure on Large-Scale Process Plants—A Natural Gas-to-Liquids (GTL) Case Study

“…In addition to the surrogate-assisted modeling cases mentioned above, this effective method was recently used to make self-optimizing implementation much easier. 14 A surrogate-assisted tool for fast implementation of the selfoptimizing procedure was presented recently by Lima et al 14 In this application, which was developed in Python, the Kriging method was used as a surrogate model. In addition to demonstrating the performance of the proposed software, three industrial-scale process plants were investigated.…”

“…In addition to demonstrating the performance of the proposed software, three industrial-scale process plants were investigated. 14 The use of surrogate models and metaheuristic algorithms for optimizing complex processes with a high computational load that is predicted by a surrogate model has been employed in numerous papers. In ref 15, Nandi et al performed a comparative investigation on the performance of two hybrid models of the multilayer perceptron-genetic algorithm (MLP-GA) and support vector regression-genetic algorithm (SVR-GA) for modeling and optimizing benzene isopropylation in the Hbeta catalytic process.…”

“…In addition, an innovative particle swarm optimization (PSO) algorithm was employed to optimize the parameters of LS-SVM and ANFIS. In addition to the surrogate-assisted modeling cases mentioned above, this effective method was recently used to make self-optimizing implementation much easier …”

“…A surrogate-assisted tool for fast implementation of the self-optimizing procedure was presented recently by Lima et al In this application, which was developed in Python, the Kriging method was used as a surrogate model. In addition to demonstrating the performance of the proposed software, three industrial-scale process plants were investigated …”

Soft-Sensor for Class Prediction of the Percentage of Pentanes in Butane at a Debutanizer Column

“…As observed in the state-of-the-art highlighted above, recent works rely on the context of Industry 4.0, digitising industrial processes [14,15] by proposing soft-sensors that integrate feature preprocessing and ML algorithms. Another hot topic in both industry and academia is automated Machine Learning (autoML) [16][17][18][19][20], which aims at enabling domain experts to build ML applications automatically [21]. As stated in [22], the ideal autoML approach involves data preprocessing, model generation, and model evaluation.…”