Abstract: For real complex industrial process plants with large-scale, time varying behavior, nonlinear, multi disturbance, and multi-objective, it is hard to develop practical mothematical models, and the complex plants are difficult to control automatically. Therefore, a new type of synthetic intelligent control system is proposed by integrating expert control with fuzzy neural network control, and can be used in real complex industrial process. The results of simulation and the application for industrial rotary kiln are presented to show the effectiveness of the proposed control system.