基于DPCAALSSVR的燃气轮机负荷估算方法

Gas Turbine Load Estimation Method Based on DPCA-ALSSVR

燃机负荷的实时估计能够为燃机参与调峰创造重要的前提条件,针对燃机负荷的实时测算问题,提出了一种基于动态主元分析结合自适应支持向量机的燃机负荷回归估计方法.首先应用动态主元分析方法挖掘出与燃机负荷相关的主要参数,获得代表相关性程度的量化指标关联度.在此基础上,利用自适应支持向量机建立燃机负荷的回归估计模型,并通过交叉验证确定模型的最优参数.采集国内某E级燃机的历史运行数据,对本文所提方法进行验证,结果表明所提方法能够对燃机负荷进行实时准确的估算,并且相比于离线建模的方式,自适应模型具有更好的泛化更新能力,为燃机的调峰奠定了可靠的基础.

Real time estimation of gas turbine load is an important prerequisite for a gas turbine to play a part in peak load regulation.To make gas turbine load real-time calculations,a load regression estimation method based on dynamic principal component analysis and adaptive support vector machine is proposed.Firstly,DPCA method is applied to mine the main parameters related to gas turbine loads,and the correlation degree which is the quantitative index representing the degree of correlation is obtained.On this basis,a regression estimation model for gas turbine loads based on adaptive least squares support vector machine is established whose optimal parameters are determined by cross validation.The proposed method is verified by collecting the historical operation data of a domestic E-class gas turbine.The results show that the proposed method can accurately estimate gas turbine loads in real time and the adaptive model is better able to generalize and update compared with offline modeling,which lays a solid foundation for peak load regulation of gas turbines.