Research On Modeling And Flexible Control Strategy For Supercritical Unit

With the vigorous development of power generation technology,supercritical unit has been the main unit of our country’s thermal power generation due to their low operating costs and high benefits.Besides,the promotion of unit parameters leads to more stringent control requirements.Thus,the introduction and development of advanced control algorithms is a general trend for the traditional control methods could no longer meet the needs of modern production.Apart from the internal demand of supercritical unit,in recent years,the problem of renewable energy accommodation caused by the continuous and rapid development of our country’s renewable energy power system has been serious.Under this circumstance,flexible operation of coal-fired power units has been an important way to assist power grid peak regulation and frequency regulation for solving before-mentioned problem.Under the above-mentioned internal demand and external background of the power system,there is an urgent need to find a better thermal power unit operation modeling scheme and control technology to lay the foundation for its safe and stable operation.Firstly,a gray box modeling method that combines mechanism analysis and data-driven intelligent identification algorithm is proposed for the modeling of supercritical unit.Based on reasonable simplifications and assumptions analysis,The nonlinear model of supercritical unit is established by analyzing the relationship between mass and energy conversion during unit operation.The steady-state parameters are obtained by regression analysis of steady-state data under typical operating conditions of the unit.In view of the actual operating data of the unit,the unknown dynamic parameters in the model are identified using immune genetic algorithm(IGA).This method can effectively deal with the complex characteristics of the unit and improve the modeling accuracy.The accuracy and effectiveness of the proposed modeling method are verified by simulation experiments relying on the on-site operating data of 1000MW supercritical unit.Secondly,to serve for the flexible operation control of the supercritical unit under a wide range of operating conditions.the model set under typical operating conditions is established for laying the foundation for the subsequent controller design.The typical operating point is selected according to the unit operating mechanism and actual operating data,and the non-linear model of the supercritical unit is linearized at the typical operating point using a small deviation linearization method to obtain a linear model set under multiple operating conditions.Then,to verify the accuracy difference of the model before and after linearization,an open-loop step disturbance experiment was carried out under the full load and steady-state conditions of the unit.Moreover,to further facilitate the design of subsequent predictive controllers.each sub-model in the linear model set obtained is discretized and converted to form a CARIMA model set.Finally,an adaptive generalized predictive control(AGPC)strategy is constructed for flexible operation of supercritical units.Based on the established CARIMA model set,the local generalized predictive control(GPC)controllers for different working conditions are designed.According to the actual operating data of the unit,the center and width of the load range corresponding to the working conditions of different models in the CARIMA model set are obtained.The weight of local GPC control output at each sample time is calculated through certain rules,and the global control law of AGPC is obtained to realize the self-adaptation of operation condition in control process.The advantages of the designed AGPC and its potential in improving the flexibility of the unit operation are verified through the tracking performance and disturbance rejective ability test.