基于PSO-SVM的传感器非线性动态补偿及LabVIEW实现  

Sensors Nonlinear Dynamic Compensation and LabVIEW Achieve Based on PSOSVM

作　　者：张霞[1] 陈保利[1] 

机构地区：[1]中北大学信息与通信工程学院

出　　处：《可编程控制器与工厂自动化（PLC FA）》2013年第9期35-38,共4页Programmable controller & Factory Automation(PLC & FA)

摘　　要：在动态测试中,由于传感器存在静态非线性失真和动态响应滞后,测量结果含有较大的误差,必须进行非线性动态补偿。实际传感器一般可以等效为一个无记忆非线性环节和一个动态线性环节串联的模型。对支持向量机进行传感器非线性动态补偿的方法进行了研究,对支持向量机核函数以及相关参数的选择进行了分析并利用粒子群优化算法进行了实现。最后在LabVIEW平台实现了补偿算法,通过仿真验证了支持向量机回归的方法在多种信噪比条件下仍然有比较好的补偿效果。In the dynamic test, because sensors exist the problems of static nonlinear distortions and dynamic response lags behind, the measurement results which have error, must carry through nonlinear dynamic compensation. The actual sensor can generally equivalent to a no memory nonlinear link and a dynamic model of linear link in series. A method for sensor nonlinear dynamic compensation using support vector machines（SVM） is studied, how to choice the kernel function of SVM and related parameters is analyzed and realized by particle swarm optimization algorithm. Finally, this compensation algorithm is achieved on LabVIEW platform, the simulation results proved that support vector machine regression method have good effects in a variety of SNR.

关 键 词：支持向量机 粒子群优化 动态补偿 LABVIEW 

分 类 号：TP212[自动化与计算机技术—检测技术与自动化装置]