Determining the reservoir characteristics is one of the important phases of reservoir study which has great importance in the simulation process. In this field, determining the static and dynamic properties corresponding to each simulation matrix block has great importance. Reducing the uncertainty in measuring and assigning static and dynamic parameters with high accuracy and low error can significantly match the simulation result with the reservoir history and its actual behavior. In the present study, the focus has been on finding a new method to facilitate the classification of reservoir rocks and increase the accuracy of measurement. At the beginning of the work, by studying the previous methods in this field and evaluating the efficiency of each method, we examined the effect of each method. Then, an attempt was made to obtain a method that is comprehensive of dynamic and static parameters. Following studies and modeling, we concluded that in order to obtain a logical and scientific classification based on reservoir physics, it is necessary to introduce two types of clustering for a reservoir. Since the movement of fluid in a porous medium has relation to factors such as porosity, permeability, viscosity, relative permeability and heterogeneity of reservoir rock, so it is necessary to introduce a method based on the classification on fluid flow and modeling. For this purpose, we used Darcy's main relationship, other relationships derived from the Darcy relationship we introduced the parameters DIFFUSION COEFFICIENT in terms of saturation (Dsw) and TEM (True Effective Mobility). The effectiveness of these methods was evaluated and presented in the form of relevant diagrams. The diffusion coefficient drives from modeling of cocurrent and countercurrent flow in the spontaneous imbibition process in the matrix block is presented in the sources; it was introduced as a practical, important and accurate basis in the classification of reservoir rocks. Based on this, it we can say that the reality of fluid flow is practically categorized. In order to apply clustering and classification, the fuzzy logic method was used, which has appeared to be very efficient and useful due to the flexibility and capability of fuzzy logic in classifying applications. In the end, a comprehensive algorithm for classification was presented and all steps for each category of reservoir rock samples as input were fully operationalized and presented.