The design of intermodal hub networks is of paramount importance in logistic operations involving multiple transportation modes like trains and trucks. In this work, we consider two non-linear optimization models for the intermodal p-hub network design problem assuming that hubs can be located anywhere in a two-dimensional space. The first formulation focuses on minimizing congestion in transport activities, while the second formulation aims to maximize hub usage using the concept of entropy. To solve the problem, we propose two heuristic algorithms which are generalizations of well-known possibilistic clustering algorithms. These heuristics are relatively easy to implement and provide a practical way to efficiently solve problems.