摘要
由于测量条件高,环状流的截面含气率直接测量往往比较困难。软测量技术的关键在于建立优良的数学模型,在分析了微粒群优化算法(PSO)和最小支持向量回归机(LS-SVR)原理的基础上,利用粒子群算法优化最小二乘支持向量回归机参数的算法,建立了软测量模型,实现了环状流的截面含气率的软测量。实验表明:该模型泛化能力强,测试精度比较高,为环状流截面含气率的测量提供了一种新的测量途径。
The void fraction of annular flow was usually rather difficulty to be directly measured on-line because of the strict measurement conditions.The key technology of soft-sensing was to construct an appropriate mathematic model,on the basis of analyzing the particle swarm optimization（PSO） algorithm and least squares support vector regression（LSSVR）,PSO algorithm was applied to optimize the parameters of LSSVR.The PSO-LSSVR mode about soft sensing of void fraction of annular flow was constructed.The soft measurement of void fraction of annular flow was realized.The results show that the model has effective generation performance and high precision,and provids a new way to measure the void fraction of annular flow.
出处
《化工自动化及仪表》
CAS
北大核心
2010年第1期34-38,共5页
Control and Instruments in Chemical Industry
关键词
环状流
截面含气率
软测量
粒子群优化算法
最小二乘支持向量回归机
annular flow
void fraction
soft measurement
particle swarm optimization algorithm
least squares support vector regression
