Design And Reduction Of Robust Controller For Supercritical Thermal Power Unit

With the trend that supercritical thermal power generating units have gradually and significantly oriented the development of the thermal power industry in our country,the control requirement of them needs to be improved continuously.However,the feature of the units including high temperature,high pressure,nonlinear characteristic and strong coupling has caused lots of problems.As a result,further researches on the supercritical boiler-turbine coordinated control system are essential.On the basis of analysis and summary of the existing research results,this paper finishes the research of the design and reduction of robust controller for the supercritical thermal power generating units,and proposes a control system which combines the coordinated control system and the control of main steam temperature.The main research results can be summarized as follows:1)Based on the robust control theory,a multivariable robust controller is designed for the supercritical coordinated control system.According to the control goal of good step tracking,a desired closed-loop system response matrix and an output-weighting filter are used to compose the augmented plant of the system.Based on the LMI method,the problem of robust H? control can be solved,so that the controller is designed which minimizes the H? norm of the closed-loop system to ensure that the system meets the condition of robust stability.By comparison with MPC,the simulation results verify that this method can fulfill the complicated control requirement of power plant and effectively improve the control quality.2)Due to the realization difficulty of the controller with high order in DCS,the controller reduction method based on system identification theory is proposed.The controller with high order is treated as a model to be identified.Through the use of GBN signal to open-loop excite the controller,the input and output data can be collected during the identification experiment.Under the premise of small relative error,the system identification is finished to get a controller with low order.In addition,with the popularity of PID controller in power plant,the paper also puts forward a method to transform low-order controller into PI controller or to directly identify the parameter of discrete PID controller.The simulation shows that the result of controller identification is satisfactory,and the low-order controller is able to maintain the great majority of the control quality before reduction,which brings new ideas of the PID structure and parameter tuning to power plant.3)Through the analysis of the characteristics and existing control methods of main steam temperature of supercritical units,the coordinated and main steam temperature combined control system is proposed to get better control result.The physical quantity,which is coupled with main steam temperature and also related with coordinated control system,can be taken seriously in the combined control system.In this way,fully considering the influencing factors,the combined control system can realize the comprehensive control of main steam temperature.Then the framework of the design and reduction of robust controller is used to the system.The simulation results verify that the combined control system is better than the traditional one of main steam temperature,reducing the interference of coupling variables and ensuring the rapid and stable control at the same time.