摘要
针对热释电红外探测器在温度识别领域的难题,本文设计了一种基于长/中波双波段钽酸锂(LiTaO_(3))热释电多元探测器,建立了基于机器学习算法的目标温度识别模型。分析了不同黑体温度下的两波段辐射量与双波段比值的变化趋势,测试了双波段比值与黑体温度的关系以及分析了双波段比值在仿真与实测中的误差,研究了基于热释电数据的决策树、随机森林算法的最优参数选择以及所搭建的模型识别准确率。研究结果表明:基于该探测器构建的温度识别系统的识别准确率最高可大于90%,为基于热释电的目标温度识别提供了一种新的技术路径,拓宽了热释电红外探测器的应用范围。
Aiming at the difficult problem of pyroelectric infrared detectors in the field of temperature recognition,a LiTaO_(3) pyroelectric multi-element detector based on long/medium wave dual band is designed,and a target temperature recognition model based on machine learning algorithm is established.The change trend of two band radiation and double band ratio under different blackbody temperatures is analyzed,the relationship between double band ratio and blackbody temperature are tested,the errors of double band ratio in simulation and measurement are analyzed,the decision tree based on pyroelectric data,the optimal parameter selection of random forest algorithm,and the recognition accuracy of the built model are studied.The research result shows that the recognition accuracy of the temperature recognition system built based on this detector can reach above 90%,which provides a new technical path for target temperature recognition based on pyroelectric,and broadens the application range of pyroelectric infrared detectors.
作者
宋泽乾
赵泽彬
胡晨晓
吴玉航
郝昕
罗文博
SONG Zeqian;ZHAO Zebin;HU Chenxiao;WU Yuhang;HAO Xin;LUO Wenbo(School of Electronic Science and Engineering,University of Electronic Science and Technology of China,Chengdu 611731,China;Chongqing Institute of Microelectronics Industry Technology,University of Electronic Science and Technology of China,Chongqing 401331,China;Chengdu Yourui Photoelectric Technology Co Ltd,Chengdu 611731,China)
出处
《传感器与微系统》
CSCD
北大核心
2024年第10期63-65,70,共4页
Transducer and Microsystem Technologies
基金
四川省科技支撑项目(2021JDRC0023)。
关键词
双波段
热释电探测器
双波段比值
机器学习
目标识别
dual band
pyroelectric detector
dual band ratio
machine learning
target recognition
