To solve the problems existing in numerical control machining processes for difficult processed metal materials, such as titanium alloy, a control method of numerical control machining processes for difficult processed metal materials based on RBF neural network was developed. And combined with CNC machining equipment, the corresponding real-time monitoring system was established. Application effect shows that, through RBF neural network control of numerical control machining processes, the surface roughness and roundness errors of workpiece are much lower than those machined in conventional machining methods. The average error of surface roughness reduced more than 50%, while the average error of roundness reduced more than 70%.