Iterative Convergence Algorithm And Software Implementation Of The Grinding Process Sequential Modular Simulation Method

The grinding process is the most common and most important process in all concentrator, the quality of its operation is essential for improving the quality and productivity of the concentrator products. This process has synthetic complexities such as multivariable, severe coupling,, strong non-linear, production operating conditions change and many key process parameters can not be measured online. It is dramatically significant to study the control algorithms, optimization algorithms and the soft sensor techniques. To reduce the venture and cost of debugging in the industry process, it is significant and has practical value to found the simulation experiment system in the laboratory.The sequential modular approach is a by-module calculation method in process simulation, Based on the sequential modular approach of the grinding process simulation possesses advantage such as good modeling framework, convenient and flexible, reusable of the element model module. But the grinding process contains circulation loop, in the use of sequential modular approach, the variables in circulation loop cutting flow must through the iterative calculation to obtain the final converged solution. This article focuses on the direct iteration method iterative convergence algorithm and its application and implement in the sequential modular simulation of grinding process.The details are as follows:(1) The convergence of the iteration algorithm such as direct iteration method, part of the iterative method, Wegstein method, Broyden method and dominant eigenvalue method are applied to the grinding classification process one or two loops in the sequential modular convergence. And in the MATLAB, the experiment and analysis prove that the circular flow variables converge in the role of iterative convergence method. Under the same circumstances, A comparative study of five kinds of convergence of iterative method convergence accuracy and speed in grinding process.(2) In the integration object simulation platform of State Key Laboratory of Process Industry, using C#language to develop five iterative convergence algorithm module including the direct iterative method module, part of the iterative method module, Wegstein method module, Broyden method module as well as the dominant eigenvalue method module.(3) To build the entire grinding process model, including the embedding of the unit model, the development of the grinding process man-machine interface and the definition of external communication interface. Iterative convergence method module was verified by the experiment in the grinding process model, prove that the correctness and validity of the iterative convergence algorithm in the sequential modular simulation of the grinding process.