This study aims to evaluate the control quality criteria for uncertain systems with diversity of reference signals. For being easy, the control algorithm is developed and tested on a model of omnidirectional mobile robot to evaluate the performance of the proposed method. The omnidirectional mobile robot is a holonomic robot that has been widely used for surveillance, inspection and transportation tasks. The radial basis function (RBF) neural network - based adaptive sliding mode control (SMC) for each input is compared with a classical SMC on the robot model. The RBF neural networks are used to estimate the nonlinear functions in the SMC law. By online training mechanism, the SMC law can adapt to the changes of control conditions. Simulations in MATLAB/Simulink indicate that the system responses are stable without steady-state error, and the overshoots archive 0.4 (%). Results illustrate that the RBF neural networks–based adaptive SMC control is stable with diversity of inputs.