Operation Optimization Of Gas-Steam Combined Cycle Based On Support Vector Machine

With the growing deteriorate of environment, the status of gas-steam combined cycle generator unit as a kind of power technology with characteristic of high efficiency and low pollution is increasing. The research of operation performance and operation optimization has important practical significance.With S209FA gas-steam combined cycle manufactured by GE as the research object, this paper adopts the methods include thermal calculation, consumption difference analysis, Least Squares Support Vector Machine(LSSVM) and Particle Swarm Optimization(PSO) algorithm to analysis the operation performance and economy of equipments of gas-steam combined cycle, and it has achieved some useful conculsions, which while provide theoretical and numerical support for mastering the actual operation performance and adjusting the operation mode of the gas-steam combined cycle generator unit. This study carried out the following areas of research work:Created analysis model in economy of S209FA gas-steam combined cycle based on thermodynamic principle, completed the optimization calculation of the operation parameters of the bottom cycle, meanwhile, analyzed the interaction of the operating parameters of the bottoming cycle, which while provide foundation for the data modeling.Consumption difference analysis is a effective means of supervising the operation economy of generator unit. This article used thermodynamic method and single factor deviation method to calculate the consumption difference of relative parameters of gas-steam combined cycle generator unit, which provide the convenience for the specific analysis of reason that affect the economy of generator unit and also indicate directions for operating crew to adjust operation of generator unit.Building off-design model of gas-steam combined cycle based on the design performance correction curves and then carried out numerical simulation, the result demonstated that the model based on the design performance correction curves can not satisfied the practical engineering requirenment. Optimizing the structural parameters of LSSVM by the Improved Chaotic Particle Swarm Optimization(ICPSO) algorithm. By the ICPSO-LSSVM, this paper built the off-design model of gas turbine, HRSG and steam turbine, and then analyzed the off-design characteristics of the equipments.Based on the ICPSO-LSSVM off-design models of the gas turbine, HRSG and steam turbine, the article built the ICPSO-LSSVM off-design models of the whole gas-steam combined cycle. With the heat consumption rate as the optimized target and using ICPSO to realize the optimal distribution of the load of the two gas turbine of the S209FA gas-steam combined cycle generator unit. The optimization results have some reference value and provide guidance for the operation of gas-steam combined cycle.