I am using peng-robinson method for prediction of VLE. How to increase the goodness of fit? Is it by just giving initial values of binary interaction parameter and checking the standard deviation??
A
ِDear Mohd Belal Haider
Did you work with an equation of state?
The binary interaction parameters between components(A and B), calculate by fitting the equation of state on the experimental data of A and B. The binary interaction coefficients between A and B evaluate by minimizing the absolute average error between calculated bubble point pressure (or dew) and experimental data.
Eslamimanesh et al lists some of the optimization methods used to obtain the parameters of thermodynamic models. You can use Simulated Annealing method (& Fminsearch in Toolbox MATLAB ) to minimize the objective function. The algorithm use in Simulated Annealing to find minimum of a function is a stochastic method. This method usually converges to global minimum.
I hope to be useful.
Dear Haider
There are different optimization method to obtained unknown parameters in the model. For your case, researchers have applied mostly search methods. They may are classified into two algorithm types such as direct and indirect search. A direct search method such as simplex method does not involve the partial derivatives of the function and therfore impossible to calculate the standard deviations of optimized parameters.However for indirect search method such as Newton method which is based on the Taylor method one can use a combination of jacobian and statistical methods to find the confidence of optimized parameters.
Best regards
Afkhamipour
- Badges
- Science topic
- Similar topics
- Correction