摘要
针对目前汽轮机真空系统漏气量的直接测量一直没有好的方法的问题,文中提出一种基于真空系统严密性试验,并结合BP网络的漏入真空系统空气量的软测量方法。该方法以影响真空下降速度的凝汽器各运行参数以及真空下降速度作为网络的输入,以常规测量方法得到的漏入真空系统空气量作为目标输出,实现对漏入真空系统空气量的软测量。数值仿真结果证明该方法可以较准确地实现在真空系统严密性试验中对漏入真空系统空气量的测量。
To counter the problem that there is not a good method to directly measure the quantity of air leaked into the vacuum system of steam turbine, it is proposed in this paper the soft-sensing technique for the measure of air leakage into vacuum system based on vacuum system tightness test of steam turbine and combined with BP neural networks. This method takes the rate of the vacuum down and the operating parameters of condenser as input of BP neural networks. And takes the quantity of air leaking into the vacuum system of steam turbine measured by the conventional method as output of BP neural networks. By training the BP neural networks, the soft measure method can be realized. The simulation results prove that this method can accurately measure the quantity of air leaking into the vacuum system of steam turbine.
出处
《系统仿真学报》
CAS
CSCD
2003年第3期444-446,共3页
Journal of System Simulation
基金
国电公司重点科技基金项目(SPKJ013-07)
关键词
汽轮机
真空系统严密性试验
BP网络
软测量
steam turbine
vacuum system tightness test
BP neural networks
soft-sensing technique