A Hybrid Optimization Approach to identification problem in thermodynamic models

The interaction parameter estimation problem in thermodynamic models is an important requirement and a common task in many areas of chemical engineering because these models form the basis for synthesis, design, optimization and control of process. For bad starting values the use gradient based result in local optimal solutions. To overcome this drawback, a global optimization approach, Simulated Annealing and genetic algorithm, has been coupled with a Nelder-Mead simplex method. To improve the accuracy of the interaction parameter estimate. The experimental ternary LLE data for extraction of toluene from n-heptane with aniline were considered in the NRTL and UNIQUAC activity coefficient model. In conclusion, the different obtained results of the prediction of liquid–liquid equilibrium are compared. These results were obtained to justify that the process of optimization recommended is very practical to estimate the interaction parameters of this ternary system.

• hybrid optimization approach
• genetic algorithm
• simulated annealing
• parameter estimation

• hybrid optimization approach
• genetic algorithm
• simulated annealing
• parameter estimation