论文部分内容阅读
采用多圈环形管用于气液两相流参数的测量,对环形管上升段水平方向内外侧差压波动信号进行了分析,采用无因次分析方法获得与差压波动信号均方根相关的特征量,得到了此特征量与容积含气率的关系模型,并在此基础上进行了实验。利用支持向量机优良的非线性映射和强大的泛化能力,建立了一个基于最小二乘法支持向量机的含气率软测量模型,给出了相应的系统结构和算法,针对LS-SVM方法参数选取困难的特点,采用遗传算法进行优化,以提高软测量的精度。仿真和实际运行结果表明,基于LS-SVM的气液两相流含气率软测量模型具有较高的估算精度与泛化能力,为气液两相流含气率的测量提供了一种简单、可靠的新方法。
The multi-circle annular pipe was used to measure the parameters of gas-liquid two-phase flow. The differential pressure fluctuation signal inside and outside of the rising section of the annular pipe was analyzed, and the characteristic related to the root mean square of the differential pressure fluctuation signal was obtained by dimensionless analysis The relationship model between this feature quantity and the volumetric gas content was obtained. Based on this, experiments were carried out. Based on the excellent nonlinear mapping and powerful generalization ability of support vector machine, a gas-bearing soft-sensing model based on least square support vector machine is established. The corresponding system structure and algorithm are given. According to the LS-SVM method parameters Select the characteristics of the difficulties, the use of genetic algorithms to optimize, in order to improve the accuracy of soft-sensing. The results of simulation and actual operation show that the gas-liquid two-phase flow soft sensing model based on LS-SVM has higher estimation accuracy and generalization ability, and provides a simple method for gas-liquid two-phase flow measurement , A reliable new method.