Dynamic simulation-based intelligent startup instruction in chemical processes

Startup is a high risk procedure in chemical industry, during which expert system is an effective way to intelligently monitor and control abnormal state. This paper proposed an intelligent startup instruction method, which uses expert system to guide operation steps and trigger first principles model to optimize startup parameters. Real-time data are converted into facts and pushed into database to produce proper instructions through inference engine. Then fuzzy logic and dynamic mechanism model are integrated into rules to quantify conclusions. The proposed method is applied to an ammonia synthesis plant simulator. Case studies demonstrate that the intelligent instruction structure can guide startup process smoothly, give proper suggestions for improper operations timely, and optimize feed valve openings for synthesis converter accurately.