The Research Of Modeling And Optimization Control In NdFeB Hydrogen Decrepitation Process

Hydrogen Decrepitation process is one of key technology in the production of high performance NdFeB magnet. The automatic control level of NdFeB hydrogen decrepitation process is limited to the process of NdFeB hydrogen crushing, which has the features that nonlinear, coupling and the important process parameters can not be online testing. At present, this process can only rely on the experience of production workers, and the process parameters can not be adjusted online. Therefore, in order to ensure that the NdFeB alloy is completely crushed, we extend the time of the hydrogen-absorbing, these results in the extension of production cycle, waste of resources and high production costs. From the angle of interdisciplinary between control and materials, the modeling and optimization of control in NdFeB hydrogen decrepitation process are studied in this thesis, and the study also designs an automatic control system which can dynamic optimize in real time.The paper deeply analysis the hydrogen absorption reaction mechanism of NdFeB, from the chemical reaction kinetics and thermodynamics, material balance and energy balance, and diffusion theory to study the effects of hydrogen absorption process, such as temperature, pressure and the flow of hydrogen in reaction. Dynamic mechanism model of the hydrogen absorption process is established by using the state space equation. Through the simulation experiments of the model, we can verify the correctness and rationality of the model. Considering that there are some hypothesis and simplification in dynamic mechanism modeling, dynamic mechanism model may not be so accurate. In view of each scientific research and industrial production process to retain a large number of data. In view of each scientific research and industrial production process to retain a large number of data, The RBF neural network prediction model is established by the technology of rolling optimization of process parameters using the data. Combination of the two models, we can proposed, multi-model optimize control strategy based on the diagnosis of furnace conditions. Through the comparison of real-time hydrogen capacity and target hydrogen capacity in hydrogen decrepitation to diagnosis current furnace hydrogen absorption, then allocate weighting factor about two models in the Optimization control algorithm through fuzzy coordinate calculations based on the current information furnace conditions, Optimization technology curve was constructed by the Optimization control algorithm. This algorithm combines the advantages of both, shielding their disadvantages. Hydrogen decrepitation automatic control system is divided into upper machine part and lower machine part, the core controller of lower machine is PLC, which combined with on-site implementation structure to complete local control. The upper machine completes kingview software remote monitoring system, the realization of the parameter optimization model of MATLAB, SQL Server database data preservation and management, as well as the configuration and the communication of MATLAB and SQL Server.The optimization control algorithm proposed by this study in the research of hydrogen decrepitation process is an exploratory work. The automatic control system proposed process control parameters online real-time dynamic optimization, and it had a very important scientific significance, research and application value.