Abstarct: Inverted pendulum know as classic problem in dynamic and control that almost used to satisfying control theorem. In recent year the Balancing robot idea proposed subject to inverted pendulum. Robot are Nonlinear, unstable and Nonholonomic so its need a complex calculation for modelling and control Published article studied Robot in different view. Some of paper define it as vehicle, some other try to modelling and so many used it to determine the control system. Different Nonlinear method used to balance, speed and trajectory control of robot also intelligence control method growth and developed such that robust and adaptive control was suggested regardless dynamic model in past decades. This thesis discuss about little balancing robot named as<Zohal> and proposed the mixture of classic and intelligence nonlinear method to control the robot. It should be note that Zohal introduced as tiny Robot not a vehicle. Experimental result show the power of Sliding-mod as a nonlinear and model baxsed method to defend disturbance but it had a large scale fluctuation cause the big trajectory errer.so we have designed a neural fuzzy network to control both of balancing and trajectory, however the good control performance and low fluctuation achieve but it hadn’t ability to deal with disturbance; because the time require for trained will not provide so imbalance is cussed by smallness of robot. To resolved this problem a linear combination of sliding mode and Neural-Fuzzy network used that combine the outputs of both controller baxsed on the error’s co-efficient, it’s mean at a same time percent of each control signal applied to the output due to the goodness of effect.