While safety is a critical criterion of the autonomous wheelchair and attracted much attention over the last few decades, the user comfort has not been considered as an important factor yet. Therefore, this article proposes a path following controller for an electric powered wheelchair (EPW) which aims to assure the comfort and safety of users during the navigation tasks. This controller is a combination of transverse feedback linearization (TFL) and model predictive control (MPC) techniques. The main contribution of this work is to consider the jerks and acceleration of EPW, directly linked to the user's relaxation, as constraints in solving an optimal control problem to design MPC controller. TFL allows to achieve path invariance properties, which is useful for path following task. The simulations are conducted in both circular path and arbitrary path in a typical office environment. Simulation results illustrate the perspective of the proposed approach where the desired path is well followed and the constraints on user's comfort are respected.