Premium
A novel fast dynamic optimization approach for complex multivariable chemical process systems
Author(s) -
Liu Ping,
Li Guodong,
Liu Xinggao,
Zhang Zeyin
Publication year - 2016
Publication title -
the canadian journal of chemical engineering
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.404
H-Index - 67
eISSN - 1939-019X
pISSN - 0008-4034
DOI - 10.1002/cjce.22633
Subject(s) - multivariable calculus , discretization , parameterized complexity , computation , process (computing) , mathematical optimization , computer science , control theory (sociology) , optimization problem , mathematics , algorithm , control (management) , control engineering , engineering , mathematical analysis , artificial intelligence , operating system
A novel fast dynamic optimization approach is proposed for complex multivariable chemical process systems, where the control variables are parameterized with different non‐uniform time grids and are treated as optimal decision variables to obtain the optimal switching structures for tackling the shortcoming caused by the conventional uniform parameterization method. Meanwhile, an adaptive fast calculation approach is proposed to calculate the differential equations so as to decrease the optimization time. The gradient formulae of decision variables are therefore further derived so that the conventional gradient‐based optimization algorithm can be utilized easily. Two well‐known complex multivariable systems in engineering are tested as illustrations and are compared with other literature reports in detail, where the uniform discretization control vector parameterization (ud‐CVP) method is also developed as the comparative base. Numerical results show that the proposed method can achieve better optimization results with fewer parameters and lower computation costs.