| It is an important method for science production organization of steelmaking, which control converter blowing process by computer. It can standard the operation of converter blowing, against the blindness, thereby improve the quality of molten steel , reduce cost of steelmaking, increase the end-point hit rate. The research improved a model of static control for converter smelting operation based on neural network .The model by the basic theory of converter process ,choose the Back Propagation(BP) network as primary form, and improved the BP algorithm .Matrix of power value of network is fixed after learning 300samples. And establish a data file which capacity is 100 samples for the real-time learning of network, in order to it's adapt of the environmental impact. it shorten training time and increased prediction accuracy after the improved. But also the predict model connect with charging system, which make it became reality that computer autocontrol charge process according to the material quantity determined by the model. And the model get a good result after it was putted into practice in converter mill of XG.In converter steel making process, the lance position directly relates to slag forming,decarbonization of molten iron ,raise of temperature and stability of process, it is very important to control the lance position rational. For that, this paper by collected the field operational data, exploited reference curve model by VB6.0. The data of in-out water temperature difference of lance can indirect reaction the converter's procedural temperature, drawing curve based on these data which convenient for control procedural temperature arise smoothly, then correct reach terminal target temperature ultimately.During design of the software, interface design is paid attention to concerning the engineering psychology. By the strong multimedia foundation of Visual Basic, a mutual software system of complete Chinese with sound , motivation, pictures and characters has been set up.
|
PDF Full Text Request