基于SSA-SVR的光纤压力传感器温度补偿研究  

Research on Temperature Compensation for Optical Fiber Pressure Sensor Based on SSA-SVR

作　　者：段松凯[1] 刘守兵 DUAN Songkai;LIU Shoubing(School of Electrical Engineering,Henan Mechanical and Electrical Vocational College,Zhengzhou He’nan 451191,China;School of Electrical and Information Engineering,Henan University of Engineering,Zhengzhou He’nan 451191,China)

机构地区：[1]河南机电职业学院电气工程学院,河南郑州451191 [2]河南工程学院电气信息学院,河南郑州451191

出　　处：《电子器件》2023年第5期1268-1274,共7页Chinese Journal of Electron Devices

基　　金：河南省高等学校重点科研项目(22A460012)。

摘　　要：针对光纤压力传感器存在的温度漂移问题,提出了基于麻雀搜索算法优化的支持向量回归(SSA-SVR)的温度补偿模型。对光纤压力传感器进行标定实验,并采用温度传感器LM35实时检测其工作温度,进而建立SVR温度补偿模型。利用SSA优化SVR的超参数,以改善温度补偿模型的补偿精度。研究结果表明,利用基于SSA-SVR的温度补偿模型对光纤压力传感器进行温度补偿后,其灵敏度温度系数和相对误差系数均提高了两个数量级,而最大绝对误差和最大相对误差只有0.5004 kPa和0.2501%,达到了温度补偿的目的。Aiming at the temperature drift problem of the optical fiber pressure sensor,a temperature compensation model combining sparrow search algorithm with support vector regression(SSA-SVR)algorithm is proposed.The SSA-SVR temperature compensation model is established by calibrating the optical fiber pressure sensor and using a temperature sensor of LM35 to detect its working temperature in real time,and then the SVR temperature compensation model is established.The hyperparameters of SVR are optimized by using SSA to improve the compensation accuracy of the temperature compensation model.The research results show that after using the SSA-SVR-based temperature compensation model to perform temperature compensation on the optical fiber pressure sensor,the sensitivity temperature coefficient and relative error coefficient are improved by two orders of magnitude,while the maximum absolute error and maximum relative error are only 0.5004 kPa and 0.2501%,achieving the purpose of temperature compensation.

关 键 词：光纤压力传感器 温度补偿 麻雀搜索算法(SSA) 支持向量回归机(SVR) SSA-SVR模型 

分 类 号：TP181[自动化与计算机技术—控制理论与控制工程] TP212.12[自动化与计算机技术—控制科学与工程]