Abstract : Stabilization problem of nonholonomic mobile robots is a difficult problem to solve. Indeed, Brockett's condition being not satisfied, stabilization using a continuous, time-invariant state feedback is impossible. Many authors proposed partial solutions to ensure trajectory tracking with local validity. Here, we propose two new approaches for trajectory tracking of nonholonomic wheeled robots. The first one is based on the use of a class of nonlinear, polytopic models. The Takagi-Sugeno (TS) fuzzy modelling framework is used which is very suitable for a synthesis of PDC (Parallel Distributed Compensation) control laws. In order to filter the measurements, an observer is added to the TS fuzzy control structure and stability of the global closed loop is ensured by using the separation principle. This represents a major advantage of the proposed method. The second approach concerns the determination of a control law with output feedback in the case of delayed measurements. Indeed, in many applications of mobile robotics, treatment or information transmission are necessary to make decisions, to find a localisation or to perceive the environment. These treatments may introduce delays more or less important. Although they may affect the system performances, these delays have so far been ignored. The proposed two new approaches have been validated on the mobile robot platform at the University of Ljubljana.