Modeling And Simulation Of Gas Turbine Generator In A Gas Steam Combined Cycle Unit

In recent years,the problem of environmental pollution has become more and more serious,Gas-steam combined generating units mainly burn clean energy and natural gas,which greatly reduces the emission of pollutants,so they have developed rapidly in recent years.Gas turbine is an important part of the combined generator set,and modeling research has become a research hotspot for the development of combined generator sets.Gas turbine system is a non-linear and strong coupling system.It is difficult to establish an accurate mathematical model of the system by traditional mechanism modeling method.Therefore,in-depth study of gas turbine system modeling method based on intelligent algorithm is of great significance for gas turbine system simulation,fault diagnosis and so on.This paper mainly studies the modeling method of gas turbine part of gas-steam combined generating unit.The mechanism of compressor,combustor and turbine in the main part of gas turbine is analyzed.It is determined that the gas turbine system is a gas turbine system with fuel quantity,IGV and speed as input,turbine exhaust temperature and power as the output of the MIMO system.Multi-Output Support Vector Regression(MSVR)is suitable for small sample,non-linear,multi-input and multi-output system modeling.Combining with the actual operation data of industrial field,the multi-output SVR model of gas turbine system under all operating conditions is established.The parameters of multi-output SVR model are determined by grid search method.The multi-output SVR model is tested under several typical operating conditions respectively to verify the universality and fitting accuracy of the model.MSVR model has large error and low precision under unstable working conditions.In order to improve the accuracy of modeling,the chaotic particle swarm optimization(CPSO)is used to optimize the parameters of the MSVR model.Chaotic particle swarm optimization makes up for the shortcoming that standard particle swarm optimization is easy to fall into local optimum,the multi-condition simulation experiment of CPSO-MSVR model is carried out.The experimental results show that the accuracy of SVR model is improved by using CPSO algorithm optimization,and the established CPSO-MSVR model of gas turbine system has good versatility under all operating conditions.