基于HLO-LSSVM算法的浮子流量计非线性校正  

Nonlinear Correction of Rotameters Based on HLO-LSSVM

作　　者：王斌[1] 姚骏[1] 张玮[1,2] 

机构地区：[1]上海大学机电工程与自动化学院,上海200072 [2]上海电力建设启动调整试验所,上海200031

出　　处：《工业控制计算机》2017年第4期103-104,106,共3页Industrial Control Computer

摘　　要：针对浮子流量计非线性校正中常用的分段线性拟合和最小二乘法不足,以及最小二乘支持向量机参数难确定问题,提出了一种基于人类学习优化-最小二乘支持向量机算法(HLO-LSSVM)的仪表非线性校正方法。首先简介了最小二乘支持向量机和人类学习优化算法,接着阐述了HLO-LSSVM算法进行非线性校正的实现过程,分别采用分段线性拟合、最小二乘法和HLO-LSSVM算法对浮子流量计非线性校正问题进行了对比。结果表明,HLO-LSSVM算法具有更优的校正效果,实现简单,具有良好的应用前景。This paper proposes human learning optimization algorithm-least squares support vector machine（HLO-LSSVM） to overcome the disadvantages of piecewise linear approximation and least squares method and the problem that it is diffi- cult to determine the parameters for least squares support vector machine for rotameters.Firstly,the theory of nonlinear cor- rection of least squares support vector machine and human learning optimization algorithm,then the process of optimization parameters of least squares support vector machine are introduced.At Last,the methods of piecewise linear approximation, least squares method and HLO-LSSVM are applied to rotameters respectively.

关 键 词：人类学习优化算法 最小二乘支持向量机 浮子流量计 非线性校正 

分 类 号：TH814.7[机械工程—仪器科学与技术] TP18[机械工程—精密仪器及机械]