Font Size: a A A

Research And Application Of Intelligent Algorithms For Initial Pressure Optimization Of Steam Turbine In Coal-fired Power Plant

Posted on:2015-06-22Degree:MasterType:Thesis
Country:ChinaCandidate:X L DuanFull Text:PDF
GTID:2272330422470697Subject:Control theory and control engineering
Abstract/Summary:PDF Full Text Request
As the development of the national economy and the adjustment of the powerconsumption structure, a growing number of thermal power units are taking part inadjusting the peak output. However, due to the efficiency of the generator unit fell sharplyfor its long-running under the off-rated load,it is necessary to optimize the operation ofthe thermal power unit. Steam turbines are important parts of thermal power units, whoseoptimal adjustments have a great significance for saving energy and economical operation.It becomes an important research direction to optimize the main pressure of steam turbines.This paper takes a600MW supercritical steam turbine unit as the research object. Thehistorical data are collected and analyzed by artificial intelligence technology, which candetermine the optimal initial pressure. Firstly, for the shortcomings of Biogeography-basedOptimization (BBO), this paper proposes an improved biogeography-based optimizationalgorithm called B-BBO-SA, in which a hybrid migration operator and simulatedannealing are introduced. The hybrid migration operator can ensure the diversity ofpopulation during the process of evolution, after which the simulated annealing is adoptedto enhance the local search ability of the algorithm. The B-BBO-SA has been proved to bereliable and effective by13classical test functions. Then the improved algorithm isadopted to optimize the parameter of relevance vector machine(RVM), which is used toset up the heat rate prediction model (B-BBO-SA-RVM). After the model is built, theB-BBO-SA can be used to search the optimal main steam pressure under certain loads.This paper adopts the historical data of a steam turbine to conduct simulationexperiment. The results show that the established model has good nonlinear fitting abilityand generalization ability, and the optimal initial pressure is obtained based on the model.Therefore, the optimal initial pressure operating curve can be drawn after the experiment,which could provide more practical guidance to the steam turbine operation.
Keywords/Search Tags:steam turbine, optimal initial pressure, heat loss rate, relevance vectormachine, biogeography-based optimization algorithm, modeling
PDF Full Text Request
Related items