基于SCA-LSSVM的电涡流传感器温度补偿方法研究  

Research on Temperature Compensation Method of Eddy Current Sensor Based on SCA-LSSVM

作　　者：李玉军[1,2] 王琛琛 焦尚彬 张青[1,2] 王庆 LI Yujun;WANG Chenchen;JIAO Shangbin;ZHANG Qing;WANG Qing(School of Automation and Information Engineering,Xi'an University of Technology,Xi'an Shaanxi 710048,China;Shaanxi Province Complex System Control and Intelligent Information Processing Key Laboratory,Xi'an Shaanxi 710048,China;National and Local Joint Engineering Research Center for Crystal Generation Equipment and System Integration,Xi'an University of Technology,Xi'an Shaanxi 710048,China)

机构地区：[1]西安理工大学自动化与信息工程学院,陕西西安710048 [2]陕西省复杂系统控制与智能信息处理重点实验室,陕西西安710048 [3]西安理工大学晶体生成设备及系统集成国家地方联合工程研究中心,陕西西安710048

出　　处：《传感技术学报》2022年第1期57-62,共6页Chinese Journal of Sensors and Actuators

基　　金：陕西省教育厅协同创新中心项目(20JY046);陕西省科技厅重点项目(2021GY-259)。

摘　　要：针对电涡流传感器容易受环境温度影响产生温度漂移这一现象,采用最小二乘支持向量机(LSSVM)建立回归模型,并将正余弦算法应用于回归模型的参数优化。通过正余弦算法对LSSVM的惩罚因子c和核函数参数δ进行优化选取,得到最佳的传感器回归模型,并和粒子群算法优化最小二乘支持向量机(PSO-LSSVM)进行比较。实验结果表明,SCA和PSO优化后模型计算结果均方误差分别为9.97×10^(-4)、4.39×10^(-3),模型优化耗时分别为578s、782s,传感器温度灵敏度系数分别为8.73×10^(-6)/℃、7.63×10^(-5)/℃。可以看出SCA算法在模型优化误差和优化效率方面均优于PSO算法。该方法提高了传感器的温度稳定性和系统检测精度,具有很强的实际应用价值。Aiming at the phenomenon that eddy current sensors are susceptible to temperature drift due to environmental temperature,a regression model is established by using least squares support vector machine(LSSVM),and the sine and cosine algorithm is applied to the parameter optimization of the regression model.The penalty factor and kernel function parameters of LSSVM are optimized through the sine-cosine algorithm,and the best sensor regression model is obtained,and compared with the particle swarm optimization least square support vector machine(PSO-LSSVM).The experimental results show that the mean square error of the model optimized by SCA and PSO is 9.97×10^(-4) and 4.39×10^(-3),respectively;the model optimization time is 578 s and 782 s,respectively;and the temperature sensitivity coefficient of the sensor is 8.73×10^(-6)/℃、7.63×10^(-5)/℃,respectively.It can be seen that SCA algorithm is superior to PSO algorithm in terms of error and efficiency of model optimization.This method improves the temperature stability of the sensor and the detection accuracy of the system,and has a strong practical value.

关 键 词：电涡流传感器 温度补偿 正余弦优化算法 最小二乘支持向量机 粒子群优化算法 

分 类 号：TP212[自动化与计算机技术—检测技术与自动化装置] TP181[自动化与计算机技术—控制科学与工程]