多批次测试数据建模新方法  

New modeling methods of multi-batch test data

作　　者：王鑫[1] 颜焱[1] 杨睿嫦[1] 苗恩铭[1] 

机构地区：[1]合肥工业大学仪器科学与光电工程学院,合肥230009

出　　处：《黑龙江科技学院学报》2010年第3期227-229,共3页Journal of Heilongjiang Institute of Science and Technology

基　　金：国家科技重大专项(2009ZX04014-023-02);浙江省教育厅科研项目(Y200805613)

摘　　要：为满足数学建模在传感器准确度分析上的准确性要求,提出了多批次测试数据数学建模的两种新方法,即数理统计建模法和参数平均建模法。将这两种方法运用到电涡流位移传感器的多批次测试数据建模中,与拟合的传感器准确度最小二乘法相比较,结果证明了数理统计建模法在传感器准确度分析上的准确性和参数平均建模法的简便性。In order to meet the accuracy of mathematical modeling,as required by sensor accuracy analysis,this paper proposes two new different modeling methods of multi-batch test data,namely,mathematical statistics modeling method and average parameters modeling method.The application of the methods to modeling electric eddy sensor＇s multi-batch test data,and comparison with least square method on the accuracy of fitted sensor prove the accuracy of the mathematical statistics modeling method designed for sensor accuracy analysis and the simplicity of average parameters modeling method.

关 键 词：传感器 精度 最小二乘法 数学建模 

分 类 号：TP614.3[自动化与计算机技术—控制理论与控制工程]