Dissolution may follow more than one mechanisms. One will be always major and other is minor. Major mathematical model can be considered most suitable.
I totally support the comment made by Mr. J. S. Patil. Along with that I maysuggest you to study some the theoretical aspect of the mathematical models for the dissolution first and then try to incorporate the same in the practical application. Several articles are available for this. After studying those you may be interested to know how you can do the same by using different software like PCP disso etc.
Dear Mohd, is your question related to the mathematical model itself or to the physico-chemical dissolution mechanism? You may fit several mathematical models to the same set of dissolution data and they don´t necessarily explain/describe what is the underlying dissolution mechanism.
You must take into account both the experimental data available, the theoretic aspects of your formulation release mechanism as well as the mathematical/statistical issues associated with the regression/fitting process.
Are the 3 models empirical or mechanistic? Are you willing to disclose some additional details (even if it's only the numerical data and models) so we may try to give some additional tips?