Modeling Of An Ultra-supercritical Unit And Nonlinear Predictive Control Applied To Its Coordinated Control

In recent years,ultra-supercritical units have been rapidly developed in China.Due to the high cycle thermal efficiency,low pollutant emissions,and fast rate of load response,the units have great influence on energy conservation,environmental protection and stability of power net.Advanced control of thermal process is the foundation of safe and economical operation of units,and the accurate model is the key of global optimal control.As the ultra-supercritical unit owns varying-parameters,strong coupling,nonlinearity,etc.,traditional control algorithms are difficult to implement the global optimal control.Thus,it is necessary to research the dynamic process of ultra-supercritical units deeply so as to build an accurate model which can be applied for the coordinated control system(CCS),and then the safety and efficiency of units can be improved.The main contents of this study are as follows:(1)Analyzing the reasons of drawbacks of immune genetic algorithm.The author puts forward to an improved immune genetic algorithm(IGA)by combining adaptive mutation in a range,mutation in high frequency with multipoint crossover.Experimental results of text functions demonstrate that the improved IGA has superior performance in parameter optimization.(2)Building a mechanism simulation model of ultra-supericritical units.After analyzing the coal-pulverized and boiler-turbine systems carefully,the mechanism structure of ultra-supercritical units is derived from the mass and energy conservation.Many groups of steady and dynamic parameters are obtained by nonlinear regression analysis and the improved IGA combined with running data.After the validation of open-and close-loop experiments,the results demonstrate the model built owns correct mechanism structure and excellent dynamic accuracy,and it can be applied for the research of model simulation.(3)Building a mechanism control model of ultra-supericritical units.As the above model owns complex structure that is difficult to obtain satisfactory control performance,the author adopts further simplification to build a model applied for control.The simple structure is derived from mass and energy conservation and then the steady and dynamic parameters are identified.After validation from many groups of running data over a wide range,the model proposed owns excellent dynamic accuracy and simple structure,applicable for the design of the CCS.(4)A nonlinear predictive control method based on mechanism model and the improved immune optimization algorithm is proposed.This method treats the mechanism model as predictor in control algorithm.The nonlinear rolling optimization problem is solved on-line using a specially designed binary-coded IGA combined with model predictive control(MPC),and then the optimal inputs of model at each sampling point are obtained.The method is illustrated via the simulating experiments of the 1000MW unit load control.The results demonstrate that the method is valid and can be referred for the design of coordinated control system of ultra-supercritical units.