Design Of Control System On The Experimental Platform For End-point Temperature Of LF Furnace

The thesis is exposed a sub-platform in "985 Project", named experimental research platform design of the control system integration and process comprehensive optimization of steelmaking-refining-casting process. The end-point temperature control system of LF furnace on the experimental platform has been completed. This experimental platform can simulate the refine production process and the refine control process in a laboratory environment, which lays the foundation of the research for production process auto-control and optimizationl.The main research contents are as follows:(1) The experimental research platform design of control system of LF furnace. Based on the simulation and control system, the network of the LF furnace and others on experimenttal platform have been realized first, and complete the configuration of PLC control syetem and WinCC interface. Then, the virtual interface and simulation can be controlled on virtual operation platform on the operation table with the linkage of virtual reality platform and simulation of LF furnace. So the basic platform of LF furnace is basically completed.(2) Based on BP neural network, the end-point temperature prediction model of LF furnace has been established. According to the law of conservation of energy, on the analysis of the factors affecting LF furnace end temperature, through the Sanming Steel Plant of the field data acquisition input and training, the model has been established..By 81 groups of data, the validation of LF furnace end-point temperature prediction is in the hit rate of 75.4%in the absolute error of ±10?.At last, the forecast platform of the end-point temperature of LF furnace has been estabilsh.(3) Research and Realization of control the end-point temperature. Based on the forecast model of end point temperature and the process of mechanism, thought a step by step control theory, in the refining stage through to the prediction of the endpoint temperature and the correction of fuzzy algorithm, endpoint temperature on the prediction and target temperature have been controled in error range. In the refining process, the detected temperature input the model to pretict the end-point temperature, and useing the fuzzy algorithm to realize the temperature deviation correction. The result is that the simulation error is within ±1 ? Finally, the interface was writed in the matlab, and the control information has been used in the control system on the experimental platform, monitoring control process at the same time.Through the study of this subject, the thesis provides the method reference of the experiment platform construction and final temperature control of the LF furnace.