BTP Prediction Based Intelligent Control System And Its Application To The Sintering Process

Iron ore sintering process is an important step in iron and steel production. As the main raw material of the blast furnace, the sinter quality has a direct influence on the yield and quality of iron and the energy consumption. At present, the backward control technology and low-level automatization of iron ore sintering process have become the bottleneck to restrict the yield and quality of the sinters in China.Since iron ore sintering is a complex industrial process with strong nonlinearity and coupling, uncertainty, time variation, and time-delay, it is difficult to meet the production requirements by using the traditional control techniques. Thus, this dissertation makes a deep research on sintering process and the related control methods and techniques by comprehensive application of multi-subject knowledge, such as the mechanism analysis, advanced control theory and artificial intelligence technique, etc. An intelligent control strategy based on the prediction for the Burning Through Point (BTP) is proposed in this dissertation, and a corresponding intelligent control system is also established, which provides a new effective approach to solve the control problems in iron ore sintering process. The main study contents and achievements include:(1) Structure of intelligent control system based on BTP predictionView of the characteristics and the main existing control problems in the sintering process, a structure of intelligent control system based on BTP prediction is presented by integrating various intelligent control methods, such as neural networks, fuzzy control and expert control. The designed control structure provides an effective solution to the control problems in sintering process.(2) Prediction model of BTPDue to the characteristics of long time-delay, nonlinearity and incompletion of parameter information in sintering process, the gray BP neural networks prediction model of BTP is established by using the grey theory and improved neural networks combinatively. When the sintering operating condition is stable, this model can predict the BTP effectively. In order to solve the insufficiency of downtrend of predictive precision when the state of sintering process is unstable, the prediction model based on PCA has been put forward to modify the gray BP neural networks prediction model. The results of simulations and actual applications show that the prediction model of BTP has high prediction precision.(3) Hybrid fuzzy-predictive control model of BTPConsidering the hybrid characters of sintering process, the hybrid fuzzy-predictive control model of BTP is established based on intelligent control methods of fuzzy control and predictive control. While BTP is in a steady state, the fuzzy control model is mainly used in the system. Otherwise, the predictive control model is mainly used. In addition, multi-model flexible switching control technique is applied to realize the smoothly switching between the two models, which solves the optimal control problem of BTP in sintering process.(4) Intelligent coordination and optimization method based on Satisfactory Solution Principle (SSP)Mainly aimed at the problem that the pallet velocity directly affects the sinter bed position continuously, the sinter bed position expert control model is set up. Then, an intelligent coordination and optimization method based on SSP is proposed to synthetically coordinate and control the BTP and the sinter bed position. The proposed method ensures the normal state of sinter bed and reduces the fluctuation of BTP, so as to realize the optimization control of the overall sintering process.(5) Intelligent control system for sintering processBased on SIMATIC PCS 7 distributed control system, the intelligent control system frame is designed by using the multi-layer distributed software architecture. The communication between the application software and basic automation system is realized with OPC. By using VC++ 6.0, the system function modules are designed, moreover, the BTP soft-sensing model, BTP prediction model and intelligent optimization and control strategies including the hybrid fuzzy-predictive control of BTP, coordination of BTP are realized. The feasibility and effectiveness of the designed system are verified by the practical control results.The application of intelligent control strategies based on the prediction for the BTP to iron ore sintering process, improves the optimizaiotn control level of the iron ore sintering process, restrains the fluctuation of BTP effectively, and improves the yield and quality of the sinters. Moreover, it reduces the labor intensity of workers, and achieves remarkable economic benefits and social benefits. Meanwhile, a set of practical and recommendable industrialized methods is supplied for the optimization control of complex industrial processes.