Frequency Domain Aanalysis And Active Disturbance Rejection Control Of Coordinated System For Supercritical Power Unit

Supercritical thermal power unit, because of its higher efficiency and lower emissions, has gradually become the main unit of the power grid system in China. But with the increasing power needs of users in the quality and a large scale new energy combined to the power grid, such as wind energy, solar energy, and other energy with stochastic characteristics, supercritical unit often participating in peak regulation and frequency modulation has become the norm. In order to ensure the quality of power supply, the higher requirements to the control quality are put forward for coordinated system in power system integrated automation. With lack of quantitative concept, the control strategy of convenional PID for coordinated system of turbine and boiler employed more auxiliaries parts and non-theoretical skills, makes the structure of system control very complicated, and its parameters turning also depends on the experience, and at the end it cannot meet the requirements of power grid for control quality.Active disturbance rejection control technique (ADRC) applies the extended state obversor (ESO) to estimate and compensate the unknown or uncertainties total disturbances in real-time, which makes the nonlinear system into a series integrator type, and improves the control quality of the complex system. Linear Active Disturbance Rejection Control (LADRC) adopts the linear feedback control structure and linear ESO, which makes the theory analysis of ADRC further developped, and its parameters turning are greatly simplified and easy to practical application. A large number of application researches show that LADRC has a stronger control ability to complex nonlinear uncertain objects.Therefore, LADRC is applied to the coordinated system of a supercritical unit in this paper, and the logic control structure and the dynamic characteristics of the plant are studied. The improved genetic algorithm is developed to identify system models. The identified models are analyzed in frequency domain, and the static decoupling design based on INA and pseudo-diagonalization is executed for the critically coupled models. Based on this work, active disturbance rejection control strategy is developed for improving the control quality of supercritical unit coordinated system, and its simulation study goes further to carry out in this paper.The main work of the thesis is following:1. The logic control structure of the coordinated system in a supercritical power unit is analyzed and simplified. With the help of the full scope simulator of the units, step disturbance tests of multi variables in coordinated system are conducted for theory studies and dynamic characteristics analysis, and the adopted complicated nonlinear controller of the system are given and analyzed.2.The processing data of the unit are acquired by MATLAB communicated with full scope simulator. Improved genetic algorithm is developed to identify the model from the data deprived at different load points of the coordinated system; and then frequency domain analysis of the model is performed by using the inverse Nyquist array method. Static decoupling design of the system is carried out based on the pseudo diagonalization.3. The structure of linear active disturbance rejection control is analyzed. The law of parameters tuning of LADRC are studied. Combined with the pseudo diagonalization decoupling in linear ADRC, the coordinated system models and controller of supercritical unit are built in Simulink. The disturbance experiments of load and main steam pressure are designed. The simulation results show that the proposed LADRC strategy based on pseudo diagonalization decoupling is obviously superior to that of the PID control strategy.4. In the quasi field enviroment of Star-90 simulator, three kinds of test strategies of the composited LADRC of the coordinated system in supercritical unit are designed: the first one is the mixed programm using Star-90 and Matlab, which is convient to observe the values of parameters in controller, save them, query and call for the historical data. This work can lay the foundation for the actural application in future; The second is using the own module in Star-90 simulation platform to build the control logic diagram, which is closer the DCS system in power plant; The third is that, in the Star-90 simulation platform, ESO of LADRC is programmed by C++and embedded to the simulation machine. The control logic diagram is built by using the modules in Star-90 and the designed ESO. This strategy can effectively shorten the machine time and improve the performance of control system. Disturbance tests of above these methods are conducted respectively. The tuning rule of the controllers parameters is given The validity of the controller and the parameters tuning is verified, which lays a good foundation for the application of LADRC in the actual power plant control.