Steam Turbine Heat Rate Prediction And Initial Steam Pressure Optimization Based On Least Square Support Vector Machine

Thermal power generation plays an absolute leading role in China’s power generation structure,and coal is the main power generation task.Improving coal energy utilization rate and reducing energy cost are the key forces for China to achieve the goal of "double carbon".Steam turbine is an indispensable device in the operation of thermal power units.Reducing the heat consumption rate of steam turbine is the key to energy saving and consumption reduction of thermal power units.However,large steam turbines often operate under variable load conditions,which will reduce the thermal economy of the unit.Therefore,optimizing steam turbine operation,saving energy and reducing consumption,improving energy utilization rate and thermal economy of the unit have become the most urgent problems faced by thermal power enterprises.In order to optimize the steam turbine operation,it is necessary to adjust the main steam pressure value to reduce the steam turbine heat consumption,so as to achieve the optimal operation state of the unit.Therefore,the prediction of steam turbine heat consumption rate and optimization of initial pressure in operation are the research objectives of this paper.The main research contents of this paper are as follows:First,in view of the steam turbine heat rate is nonlinear characteristics and complicated calculation,using the test method and the theoretical calculation method to calculate the heat rate has many defects,based on Least Squares Support Vector Machine algorithm(LSSVM)heat rate prediction model is established,and by using Particle Swarm Optimization(PSO)algorithm to the model parameter optimization,The test results show that the improved PSO-LSSVM model significantly improves the prediction accuracy of heat rate.Secondly,in view of the characteristics of nonlinear,time-varying and variable operating conditions of steam turbine,this study adopts a multi-model modeling method combining Fuzzy C-Means Clustering(FCM)algorithm and Least Squares Support Vector Machine(LSSVM),and uses Particle Swarm Optimization(PSO)algorithm to search the optimal model parameters.The multi-model was applied to predict the heat rate of a steam turbine unit.By comparing the prediction curves and performance indexes of the multi-model,PSO-LSSVM single model,LSSVM and SVM models,it can be found that the prediction performance of the multi-model is better than that of LSSVM model,SVM model and PSO-LSSVM single model.This multi-model modeling method has obvious advantages in predicting the heat rate and provides an effective means for calculating the steam turbine heat rate accurately and efficiently.Finally,based on the established FCM-PSO-LSSVM multi-model of heat rate prediction,Particle Swarm Optimization(PSO)algorithm was used to search for the optimal initial operating pressure within the feasible pressure range,and the optimal initial operating pressure curve was obtained and used as the main steam pressure setting value of steam turbine operation.Compared with before optimization,the steam turbine heat rate after initial pressure optimization is significantly reduced.The results show that PSO algorithm can be used to optimize the sliding pressure curve,and the optimal initial pressure curve in operation has very strong guidance,which can optimize the steam turbine operation to achieve the purpose of saving energy and reducing consumption and improving the economic benefits of thermal power plants.