The pore structure evolution is affected by mining. This evolution influences the permeability behavior of porous media and the fluid flow path selection. An equivalent topological network model is used to describe the fluid seepage properties and path selection taken example by a complex circuit in series-parallel connection. This model is applied to the quantitative analysis of the geometric structure's influence on fluid flow and flow path selection. The connection between permeability and geometric structure parameters is discussed. The description of the fluid flow process on the mesoscale and the fluid properties on the macroscale are connected. An optimization algorithm and topological network model are discussed. The equivalent network model and the optimization algorithm to analyze the flow process and the flow quantity distribution are used. In order to check the algorithm, a numerical simulation of seepage in porous media is done by COMSOL Multiphysics software. The results of the topological network model and the COMSOL Multiphysics simulation are matched well. In that case, under the circumstances that only the geometric structure of porous media is known, the equivalent topological network model and the flow path preferential selection algorithm can describe the fluid flow in porous media and the flow path selection behavior can be predicted.