A control algorithm is basically a mapping that maps the controller_ s input values into output values. This mapping can be specified by analytical equations, decision trees, fuzzy rules, neural networks, or a combination. The design of the neural controller is carried out by identified the motor _ antenna for earth station plant, The neural network enhancement the PID control, after that and by designing a neural controller and testing the performance of the system it is found that the neural controller satisfy the steady state error requirement of the system but doesn't ability to deal with the uncertainty so the fuzzy system compound with neural controller both of them are model free estimators and share the common ability to deal with uncertainties and noise. In this thesis, Neuro-Fuzzy positioning system for a large Antenna is conducted to improve accuracy of azimuth positioning system. Reliability and accuracy of the system mainly depends on numbers of learning points, and sufficient signal strength samples to represent each point with least error are presented. A Neuro-Fuzzy controller is developed and simulated on our system by using Matlab program. Neuro-Fuzzy performance is compared with a neural controller. The former gave better performance. The controller performance can still be improved by training the Neuro-Fuzzy network with more number of rules and input-output combinations.