基于MGF-SR-GM(1,1)混合模型的血压模拟器误差修正方法研究  

Research on Error Correction Method of NIBP Simulators Based on MGF-SR-GM(1,1)Hybrid Model

作　　者：洪程鑫 陆金霞 杨银银 陆莉 郭亚琴 秦朗 Hong Chenxin;Lu Jinxia;Yang Yinyin;Lu Li;Guo Yaqin;Qin Lang(Nangtong Institute of Metrology and Measurement,Nangtong 226011)

机构地区：[1]南通市计量检定测试所,南通市226011

出　　处：《仪器仪表标准化与计量》2025年第2期37-40,共4页Instrument Standardization & Metrology

摘　　要：血压测量的准确性对于诊断和监测各种医疗状况至关重要,研究血压模拟器的误差信息,对于提高血压模拟器的准确性和可靠性具有重要的意义。本文根据血压模拟器的误差数据拓展成多维的矩阵,结合逐步回归的理论,建立起均生逐步回归模型(MGF-SR)拟合实际误差,截取尾部的残差数据建立起GM(1,1)修正模型来提高预测的精度,形成MGF-SR混合模型。模拟实验证明了该模型的有效性,为后续血压测量的潜在误差提供了测试办法和判断依据。The accuracy of blood pressure measurement is crucial for diagnosing and monitoring various medical conditions.Studying the error information of blood pressure simulators has great significance for improving the accuracy and reliability of blood pressure simulators.The paper expands the error data of the blood pressure simulator into a multidimensional matrix,and combines the theory of stepwise regression to establish a Mean Generation Function Stepwise Regression(MGF-SR)model to fit the actual error.To improve the accuracy of prediction,a GM(1,1)correction model was established by truncating the residual data at the tail end.Finally,a MGF-SR hybrid model is formed.The effectiveness of the model has been demonstrated through simulation experiments,providing a testing method and judgment basis for potential errors in subsequent blood pressure measurements.

关 键 词：血压模拟器 精度 准确性 回归模型 

分 类 号：R318.6[医药卫生—生物医学工程]