Documents disponibles écrits par cet auteur

Mnemonic enhancement optimization (MEO) for real-time optimization of industrial processes / Xueyi Fang in Industrial & engineering chemistry research, Vol. 48 N°1 (Janvier 2009) 

Public ISBD [article]  in Industrial & engineering chemistry research > Vol. 48 N°1 (Janvier 2009) . - P. 499-509 Titre : Mnemonic enhancement optimization (MEO) for real-time optimization of industrial processes Type de document : texte imprimé Auteurs : Xueyi Fang, Editeur scientifique ; Zhijiang Shao, Editeur scientifique ; Zhiqiang Wang, Editeur scientifique Année de publication : 2009 Article en page(s) : P. 499-509 Note générale : Chemical engineering Langues : Anglais (eng) Mots-clés : Mnemonic  Enhancement  Optimization  (MEO)  Optimization  of  Industrial  Real-time  optimization  (RTO) Résumé : In this paper, the model-based real-time optimization (RTO) is viewed as a kind of nonlinear parametric optimization problem which is solved repeatedly when parameter values change. A novel RTO strategy—mnemonic enhancement optimization (MEO)—is proposed. The method preserves the past optimal solutions and corresponding parameter values as experience and approximates the optimum based on the experience. The approximation is used by the optimization algorithm as a starting point to find the real optimum. The optimum is proved to be a continuous function of the parameter. This ensures that the distance between the optimum and the initial point tends to decrease as RTO continues to run. Thus MEO can improve the performance of RTO continually. Numerical experiments illustrate the continuity of the optimal set mapping, and the MEO method is compared with the traditional method. The results show that MEO outperforms the traditional method concerning the solution time, the number of iterations, and the percentage of successful optimizations. En ligne : http://pubs.acs.org/doi/abs/10.1021/ie800166p [article] Mnemonic enhancement optimization (MEO) for real-time optimization of industrial processes [texte imprimé] / Xueyi Fang, Editeur scientifique ; Zhijiang Shao, Editeur scientifique ; Zhiqiang Wang, Editeur scientifique . - 2009 . - P. 499-509. Chemical engineering Langues : Anglais (eng) in Industrial & engineering chemistry research > Vol. 48 N°1 (Janvier 2009) . - P. 499-509 Mots-clés : Mnemonic  Enhancement  Optimization  (MEO)  Optimization  of  Industrial  Real-time  optimization  (RTO) Résumé : In this paper, the model-based real-time optimization (RTO) is viewed as a kind of nonlinear parametric optimization problem which is solved repeatedly when parameter values change. A novel RTO strategy—mnemonic enhancement optimization (MEO)—is proposed. The method preserves the past optimal solutions and corresponding parameter values as experience and approximates the optimum based on the experience. The approximation is used by the optimization algorithm as a starting point to find the real optimum. The optimum is proved to be a continuous function of the parameter. This ensures that the distance between the optimum and the initial point tends to decrease as RTO continues to run. Thus MEO can improve the performance of RTO continually. Numerical experiments illustrate the continuity of the optimal set mapping, and the MEO method is compared with the traditional method. The results show that MEO outperforms the traditional method concerning the solution time, the number of iterations, and the percentage of successful optimizations. En ligne : http://pubs.acs.org/doi/abs/10.1021/ie800166p

Random sampling-based automatic parameter tuning for nonlinear programming solvers / Weifeng Chen in Industrial & engineering chemistry research, Vol. 50 N° 7 (Avril 2011) 

Public ISBD [article]  in Industrial & engineering chemistry research > Vol. 50 N° 7 (Avril 2011) . - pp. 3907–3918 Titre : Random sampling-based automatic parameter tuning for nonlinear programming solvers Type de document : texte imprimé Auteurs : Weifeng Chen, Auteur ; Zhijiang Shao, Auteur ; Kexin Wang, Auteur Année de publication : 2011 Article en page(s) : pp. 3907–3918 Note générale : Chimie industrielle Langues : Anglais (eng) Mots-clés : Nonlinear  programming  solver Résumé : Nonlinear programming solvers play important roles in process systems engineering. The performance of a nonlinear programming solver is influenced significantly by values of the solver parameters. Hence, tuning these parameters can enhance the performance of the nonlinear programming solver, especially for hard problems and time-critical applications like real time optimization and nonlinear model predictive control. Random sampling (RS) algorithm is utilized to tune the nonlinear programming solver for solving hard problems. By introducing an iterated search technique, heuristic rules and advanced termination criteria, an enhanced random sampling algorithm is developed to determine parameter configuration that work significantly better than the default. These random sampling-based methods can handle all kinds of parameters (e.g., categorical, integer, and continuous) of the nonlinear programming solver. Numerical results with parameter configurations from the proposed random sampling-based methods show remarkable performance improvement compared with the defaults. Future works toward the development and improvement of the automatic parameter-tuning tool for the nonlinear programming solvers are also outlined. DEWEY : 660 ISSN : 0888-5885 En ligne : http://pubs.acs.org/doi/abs/10.1021/ie100826y [article] Random sampling-based automatic parameter tuning for nonlinear programming solvers [texte imprimé] / Weifeng Chen, Auteur ; Zhijiang Shao, Auteur ; Kexin Wang, Auteur . - 2011 . - pp. 3907–3918. Chimie industrielle Langues : Anglais (eng) in Industrial & engineering chemistry research > Vol. 50 N° 7 (Avril 2011) . - pp. 3907–3918 Mots-clés : Nonlinear  programming  solver Résumé : Nonlinear programming solvers play important roles in process systems engineering. The performance of a nonlinear programming solver is influenced significantly by values of the solver parameters. Hence, tuning these parameters can enhance the performance of the nonlinear programming solver, especially for hard problems and time-critical applications like real time optimization and nonlinear model predictive control. Random sampling (RS) algorithm is utilized to tune the nonlinear programming solver for solving hard problems. By introducing an iterated search technique, heuristic rules and advanced termination criteria, an enhanced random sampling algorithm is developed to determine parameter configuration that work significantly better than the default. These random sampling-based methods can handle all kinds of parameters (e.g., categorical, integer, and continuous) of the nonlinear programming solver. Numerical results with parameter configurations from the proposed random sampling-based methods show remarkable performance improvement compared with the defaults. Future works toward the development and improvement of the automatic parameter-tuning tool for the nonlinear programming solvers are also outlined. DEWEY : 660 ISSN : 0888-5885 En ligne : http://pubs.acs.org/doi/abs/10.1021/ie100826y