A new model-based predictive control algorithm for vehicle trajectory control is proposed by using vehicle velocity and sideslip angle. Based on the error function combined with vehicle velocity and side slip of a bicycle model, a predictive control method has been proven to be useful on low velocity. Thus, it could be applied for an autonomous vehicle without a driver. Although an autonomous robot is not necessary to be driven with a high velocity, a commercial vehicle has to be driven at high velocity. Thus the previous predictive control formulation is not enough for a commercial driving system. This study is proposed to enhance the capacity of the predictive controller for rather high speed vehicles.