Process optimization of reactives systems by partially reduced SQP methods

Abstract

The increasing demand for optimization in process control of reactive systems necessitates the development of fast and reliable software for the numerical computation of optimal controls taking into account the special structures of the chemical systems under investigation. A recently developed algorithm based on partially reduced SQP methods is used for the computation of optimal controls following the direct approach. In this work, this algorithm is coupled with a package for the simulation of homogeneous reactive systems, which provides the specific routines for the description of the problem at hand. The combined code takes advantage from the fact that the number of influence variables to be determined is rather low compared with the overall number of variables. For example, the computation tool developed is applied to study the homogenous oxidative coupling of methane.

Catalytic reforming of logistic fuels at high-temperatures

“…Recently, a new mathematical algorithm was developed to not only optimize the operating conditions but also the catalyst loading. [81][82][83] These new computational tools were recently applied to optimize catalytic oxy-dehydrogenation of ethane at high temperatures and short contact times by Minh et al 81 In their study, radical interactions in gas and surface chemistry were shown to play a decisive role for yield increase. These tools may support design and operation of catalytic fuel reformers in the near future.…”

Aerodynamic shape optimization using simultaneous pseudo-timestepping

“…A detailed discussion of this approach can be found in [34,35]. Industrial applications of the RSQP concepts are discussed in [2,26,36,37]. Here we present the outline of the method.…”

Data-driven model reduction-based nonlinear MPC for large-scale distributed parameter systems

“…Although not all control problems are constrained, many do require the satisfaction of nonlinear inequalities. The need to tackle inequalities is accommodated in purpose-designed algorithms for predictive and optimal control [18][19]. Such constraints could express design constraints for process operation [20], economic/environmental [21] or stability conditions [22].…”