Visual Simulation Research On Flatness Cloud Inference Control For Cold Strip Mill

In the world economy, plate and strip plays an important and extensive role. It is oneof the main steel products. With the progress of society and the development of rollingtechnology, people put forward new requirements for plate and strip’s quality. Flatness isone of the most important indicators of the quality of the test plate and strip. It makes thetheory of flatness recognition and control becoming a key technology in the field of rolling.Aiming at the existing many factors of flatness control, applications of intelligent controltheory offer a new way to solve the problem of flatness control. It has been widely appliedin the process of industry rolling at the same time. In this paper, the research subject ischosen as cold strip rolling mill flatness cloud inference control. The T-S cloud inferenceflatness intelligent control model based on Genetic Algorithm (GA) is designed andvisualization of the flatness recognition and control is realized.First of all, the existing flatness recognition and control theory has many problems,such as low precision, complex and changeable parameters, modeling difficulty.According to the latest advances of artificial intelligence, a new method, called cold striprolling mill flatness cloud inference control, is proposed in this paper. This method isbased on T-S cloud inference neural network. Through comparing the similarities anddifferences of cloud model and gaussian membership function, the cloud model is broughtin the T-S fuzzy neural network. And the T-S cloud inference neural network isconstructed. The new network not only has joined the uncertainty of the concept of cloudmodel, but also equips the advantages of T-S fuzzy neural network. It is a stable andeffective network.In addition, the new T-S cloud inference neural network is applied in flatnessintelligent control area. Aiming at a6-high900HC reversible cold rolling mill, T-S cloudinference neural network flatness recognition model and predictive model based on GAare developed and flatness intelligent control system is designed. By comparing thesimulation results, GA gives full play to the ability of global optimization. It improves theaccuracy of flatness recognition and control.Finally, under the condition of without increasing the hardware investment,3ds max is used for3D modeling of the rolling mill. Combined with hybrid LabVIEW andMATLAB technology, T-S cloud inference flatness intelligent control model is packagedby graphical programming language (G language) and flatness intelligent control virtualsimulation experiment system is designed. It takes the form of a kind of human-computerinteraction interface to realize the visualization of flatness pattern recognition and control.What’s more, flatness control results are showed intuitively and conveniently.