Visual Continuous Casting Break Out Prediction System

With the flourish development of the world economy and the fast movement of modern construction, the steel companies are playing more and more important roles, while the competition between them is becoming more and more intensive, thus the high technology becomes the key, the technology about breakout prediction is the most important. This dissertation bases on the primary logical breakout prediction system, it builds a new system which integrates the neural networks and imaging technology into one, analyses and studies the theory of both neural networks and imaging, basing on this, builds mathematical model, realizes a new system, the new system not only has the function of primary breakout system but also analyzes the phenomenon of breakout and the thermodynamic behavior of billet through intuitionistic planar imaging as a result, we will get a better result of breakout prediction.The test data comes from the historical data of primary logical prediction system in BAO Steel Company, using them to study and analyze, the running of system and the creation of interface are all basing on the historical data. The result of simulation proves that the new system not only has the precision of primary neural network but also study the thermodynamic behavior of billet more conveniently. It gives the scientific and practical references to improving the nicety ratio of prediction, if it can integrates the real time data, the effects will be better. The dissertation includes:First, analyses and studies the mechanism of break out;Second, the application of the theory about neural networks;Third, the application of the theory thermodynamic imaging;The last, summarizes the work of this dissertation, indicates the shortage, and gives the direction and prospect of visual breakout prediction system .