Thank you. As I know about weighted sum method, in spite of becoming a popular method, the results obtained by weighted methods on the Pareto front have serious problems with changing the coefficients. It has been noticed that changing weighting by small values may lead to remarkable changes in the objective vectors while changing weighting by large values may cause almost negligible changes in the objective vectors. This phenomenon is mainly because of the fact that this method is not a Lipschitzian function of weighting.
In addition, I have used an AHP method to select best Pareto solution after generating Pareto optimal solutions by epsilon constraint method and NBI method.
Gams source code is based (a fact that is still unacknowledged by developers of software, but acknowledged vaguely by Debru) paper with Akira Takayama : http://www.people.ku.edu/~melhodiri/Pareto-Optimality-of-Competitive-Equilibrium-in-a-General-Equilibrium-of-Economic-Growth%20.pdf
Anyway the paper accommodates multiple criteria
If anyone (including Gams developers) wishes to collaborate on developing a modefied source code to take care of multiple criteria, then I am ready. Takayama is not ready because de has decided to permanently suspend his Yoga breathing excercise.