基于PSO-SVM的电涡流传感器温度补偿的研究
|
摘要
为降低温度在电涡流传感器上对于测量数据的干扰,采用粒子群改善支持向量机(PSOSVM)曲线矫正的方法,构建了一种软件补偿模型。利用粒子群算法优化参数的功能对SVM核函数进行搜索寻优;根据向量机迫近非线性函数的优点,完成电涡流传感器的温度补偿。数据结果表明,灵敏度温度系数由1.35×10~(-2)/℃提升到5.59×10~(-4)/℃,优化效果提升了1个精度;零位温度系数由1.35×10~(-2)/℃提升到1.81×10~(-4)/℃,优化效果提升了近2个精度,很好地消除了非目标参数对电涡流传感器的影响。
In order to reduce the interference of temperature on the eddy current sensor,a software compensation model is constructed. The paper presents a nonlinear correction model based on particle swarm optimization and support vector machine(PSO-SVM). The model firstly uses the particle swarm to search and optimize the kernel function and the parameter of the SVM. SVM model can approximate any arbitrary nonlinear functions,so it can help the eddy current sensor achieve temperature compensation. The results show that the sensitivity temperature coefficient is raised from1.35×10~(-2)/℃ to 5.59×10~(-4)/℃ after compensation,which is improved by an order of magnitude,the zero temperature coefficient is raised from 1.35×10~(-2)/℃ to 1.81×10~(-4)/℃,which is improved by almost two orders of magnitude,and the purpose of temperature compensation is achieved successfully.
In order to reduce the interference of temperature on the eddy current sensor,a software compensation model is constructed. The paper presents a nonlinear correction model based on particle swarm optimization and support vector machine(PSO-SVM). The model firstly uses the particle swarm to search and optimize the kernel function and the parameter of the SVM. SVM model can approximate any arbitrary nonlinear functions,so it can help the eddy current sensor achieve temperature compensation. The results show that the sensitivity temperature coefficient is raised from1.35×10~(-2)/℃ to 5.59×10~(-4)/℃ after compensation,which is improved by an order of magnitude,the zero temperature coefficient is raised from 1.35×10~(-2)/℃ to 1.81×10~(-4)/℃,which is improved by almost two orders of magnitude,and the purpose of temperature compensation is achieved successfully.
引文
[1]刘君华.智能传感器系统[M].2版.西安:西安电子科技大学出版社,2010.
[2]刘焱,王烨.位移传感器的技术发展现状与发展趋势[J].自动化技术与应用,2013,32(6):76-80.
[3] Futane N P,RoyChowdhury S,RoyChaudhuri C,et al. Analog ASIC for improved temperature drift compensation of a high sensitive porous silicon pressure sensor[J].Analog Integr Circ and Sig Process,2011,67(3):383-393.
[3]卢文科,朱长纯,方建安.霍尔元件与电子检测应用电路[M].北京:中国电力出版社,2005.
[4]李如发.基于多元回归和支持向量机的电涡流传感器的研究[D].上海:东华大学,2014.
[5]陈果.基于遗传算法的支持向量机时间序列预测模型优化[J].仪器仪表学报,2006,27(9):1080-1084.
[6] Yang B,Cheng L.Study of a new optimization algorithm based on the standaed PSO[J]. Journal of Optimization Theory and Applications,2013,158(3):935-944.
[7]张妤,谢永华,穆丽,等.基于支持向量机的电容式传感器温度补偿研究[J].传感器与微系统,2009,28(6):40-42.
[8]刘毅,王海清,李平.局部最小二乘支持向量机回归在线建模方法及其在间歇过程的应用[J].化工学报,2007,58.
[9]张丽平.粒子群优化算法的理论及实践[D].杭州:浙江大学,2005.
[10]刘安丰,陈志刚,曾锋.基于PSO的无线Mesh网关优化部署算法[J].传感器学报,2008,21(7):1251-1258.
[11]张树团,张晓斌,雷涛,等.基于粒子群算法和支持向量机的故障诊断研究[J].计算机测量与控制,2008,16(11):1571-1581.
[2]刘焱,王烨.位移传感器的技术发展现状与发展趋势[J].自动化技术与应用,2013,32(6):76-80.
[3] Futane N P,RoyChowdhury S,RoyChaudhuri C,et al. Analog ASIC for improved temperature drift compensation of a high sensitive porous silicon pressure sensor[J].Analog Integr Circ and Sig Process,2011,67(3):383-393.
[3]卢文科,朱长纯,方建安.霍尔元件与电子检测应用电路[M].北京:中国电力出版社,2005.
[4]李如发.基于多元回归和支持向量机的电涡流传感器的研究[D].上海:东华大学,2014.
[5]陈果.基于遗传算法的支持向量机时间序列预测模型优化[J].仪器仪表学报,2006,27(9):1080-1084.
[6] Yang B,Cheng L.Study of a new optimization algorithm based on the standaed PSO[J]. Journal of Optimization Theory and Applications,2013,158(3):935-944.
[7]张妤,谢永华,穆丽,等.基于支持向量机的电容式传感器温度补偿研究[J].传感器与微系统,2009,28(6):40-42.
[8]刘毅,王海清,李平.局部最小二乘支持向量机回归在线建模方法及其在间歇过程的应用[J].化工学报,2007,58.
[9]张丽平.粒子群优化算法的理论及实践[D].杭州:浙江大学,2005.
[10]刘安丰,陈志刚,曾锋.基于PSO的无线Mesh网关优化部署算法[J].传感器学报,2008,21(7):1251-1258.
[11]张树团,张晓斌,雷涛,等.基于粒子群算法和支持向量机的故障诊断研究[J].计算机测量与控制,2008,16(11):1571-1581.