Objective
The project is aimed primarily at constructing a generic, intelligent, sophisticated, modular, and configurable diagnosing system that can be installed at rolling mills and similar other processes in Steel plants, with a view to improving the process control strategy. The proposed system will incorporate comprehensive mathematical models combined with neural network algorithms and Fuzzy reasoning in order to improve quality of strips and to minimize mill downtime.
With this purpose and for developing generalised strategies, three hot rolling mills having requirements of different but similar nature were identified, where only partial successes have been achieved in this direction as results of various research and scientific efforts made in the past several years. In all these localities varying number of measuring instruments (pyrometers, thickness- and width gages etc.) are installed at different stages along the production path. In some places the signals are fed to expert systems, which perform certain diagnostic and control activities to some extent, while in others they are partially employed manually at present.
Based on the know-how of the project partners in the fields of supervision, diagnosis and control and under consideration of the newest publications in these fields, practical and powerful methods and tools for control-performance monitoring and supervision will be developed and installed in at least one of the participating hot mills.