| The PID control strategy is usually adopted in the control systems in power plant atpresent. However, the process of the power plant is very complex, which involves thetransmission of the substance in the tubes, combustion of the coal in the furnace and heatexchange. Furthermore, the power of the unit is changed with the change of needed loadfrequently. So, the combustion process is a kind of nonlinear system with strongcoupling, more conditions and significant delay. The need of the control system ofcombustion process can not be satisfied by the normal control strategies. For theadvantages of simple to design, excellent robustness, and better prediction ability, theinternal model control (IMC) shows the strong vigor to solve the control problem of thethermal process with large time delay. However, the dynamic characteristics ofcontrolled processes in power plant are different with each other. Based on the status ofIMC, we have studied systematically in this dissertation the IMC method and its relatedtheories for the processes with uncertain parameters and large time delay. Our studyaddresses the following topics:Firstly, the IMC strategy for the stable linear process is introduced. We summarizethe structure, the method to design and properties of IMC. To design the control systemabilities to track the set point and reject external disturbances respectively, thetwo-freedom IMC is adopted. The research and simulation results showed that someexternal disturbances of the unstable controlled processes can not be rejected by classicalIMC. So the internal feedback controller is used to stabilize the unstable process, andthen the IMC system can be designed. However, this control system can not reject theoffset introduced by some external disturbances, a PID compensation controller isdesigned in the control system.A kind of IMC controller is designed for the stable process with large time delay,and the permitted variation ranges of controlled process parameters are given. Theadaptive IMC controller is presented for the controlled process with large time delaywhose parameters vary slowly and an improved adaptive rule is presented to identify thecontrolled process. For the unstable controlled process with large time delay, the Taylorexpansion is used to approach the delay element the process, the IMC control system isdesigned and the permitted ranges of the feedback controller parameters to keep thecontrol system stable are given.For the processes with large time delay whose parameters vary quickly, the adaptivemultiple IMC strategy is studied. To restrain the chattering reduction when the controlleris switched, the incremental controller is used and the current element controller is notbeen switched when the error is in a small range. The localization technology is used tooptimize the model set and improve the performance of the control system.For the inversible nonlinear processes with large time delay, the neural networkIMC control method is studied. Using IBP neural network training algorithm, the highercomputation speed is obtained and the local minimum problem is avoided. Simulationresults showed the effectiveness of this algorithm.Taking the superheated steam system as an example, the application of IMC for thethermal processes with large time delay is studied. Based on the analysis of the structureand the dynamic characteristics of the superheated steam system, the adaptive multipleIMC superheated steam control system is designed to make the controlled systemeffective when the parameters of the controlled process vary with the change of the unitload. Based on the characteristics of the superheated steam system, combustionfeedforward, unit load feedforward and the mill start/stop feedforward are designed toimprove the performance of the control system. The radiation energy is used as a middlesignal to represent the combustion condition. Simulation results showed the excellentperformance of the designed control system.The conclusion and perspective are given in the end of the paper.
|
PDF Full Text Request