Research On Modeling And Decoupling Control Of Ball Mill Pulverizing System

At present, ball mill with double inlets and outlets equipped with positivepressure direct blowing pulverizing system has been mainly adopted by300MWgenerating units and above. For the nonlinear process with multi-variables, strongcoupling, large time delay and numerous uncertain interference factors, it is verydifficult to establish an accurate mathematical model. As a typical three input threeoutput systems, the problems of modeling and decoupling control of ball millpulverizing system are still hot topics in today’s thermal power studies.It is concerned with two main problems in this thesis. The first one is toestablish the mathematical model of the controlled object. First of all, the maincomponents, the working principles and the performance analysis of ball mill withdouble inlets and outlets pulverizing system are introduced in detail. Then, therelated balance equations are written with the simple intuitive mechanism modelingmethod. The type of BBD4060ball mill with double inlets and outlets is chosen asan example to put the data into the equations and then to sort them, through whichthe mathematical model of the type of ball mill with double inlets and outlets can beobtained. In order to verify the accuracy of the model, the mechanism model of theball mill with double inlets and outlets is simulated with Matlab/Simulink, and thesimulation results are compared with the field measured. In addition, the step test isadded into the given input during the simulation process to solve the transferfunction of the system, which lays the foundation for the further control simulation.The following is the decoupling control of the ball mill with double inlets andoutlets after completing the modeling. Internal model control method and supportvector machine generalized inverse with internal model control method are adoptedin this thesis. For the internal model control method, the principle of internal modelcontrol of single variable is given firstly. The design of controller of internal model decoupling is introduced，and then the analysis of decoupling control is simulatedwith Matlab/Simulink, Secondly, the fundamental concepts and principles of supportvector machine generalized inverse are discussed and the algorithm of supportvector machine generalized inverse with internal model control is provided. Finallythe simulation results are compared with the internal model decoupling control,which can be known that the support vector machine generalized inverse withinternal model control method combines the advantages of two algorithms. And, theeffects are superior to the one only using internal model control method. Thismethod provides a new solution to the multivariable, strong coupling, nonlinear andcomplex controlled system.