This paper presents a novel methodology for designing a neuro-fuzzy control for controlling systems with nonlinearities of known structures and parameters. Due to the different nonlinearities inherent in the system dynamics, we propose a neural-fuzzy control that processes both numerical and linguistic information. The proposed control has some characteristics and advantages, the inputs and outputs are fuzzy numbers or numerical numbers, the weights of the proposed neuro-fuzzy control are fuzzy weights owing to the representation forms of the alpha-level sets. Different tests are performed to study the effects of different alpha-cut techniques on the closed loop system performance, the first test addresses the effect of alpha-cut techniques, the second test is related to the effect of number of labels and finally the effect of defuzzification techniques. This approach is able to process and learn numerical information as well as linguistic information. It can be used as an adaptive fuzzy controller. As a model example of nonlinear system, the DC shunt motor is considered due to its importance in electric drive. Computer simulations are included to help in conducting the study.