The flatness control system is a multivariate and complex system with uncertainty of parameters. General modeling approaches can not satisfied the high precision demand of rolling process. And it is difficult to establish a precise mathematical model of the rolling mill. In this paper, T-S cloud inference neural network is proposed. It is constructed by cloud model and T-S fuzzy neural network. The uncertainty of cloud model for processing data and the rapidity of fuzzy logic are both taken into account synthetically. And it is applied for 900HC reversible cold rolling mill. A Virtual Flatness Intelligent Control System (VFICS) is successfully developed by integrating hybrid MATLAB and LabVIEW with T-S cloud inference neural network. VFICS is a virtual experiment system, which has realized flatness pattern recognition, flatness prediction and flatness control. Genetic algorithm (GA) and particle swarm optimization (PSO) are used to train the network. Simulation results show that control effect of GA has higher accuracy and faster convergence and it is suitable for controller design. They are provided to confirm the performance and effectiveness of the proposed control method.