基于岭估计的硅酸盐混凝土抗压试验分析
|
摘要
通过测试混凝土养护28 d期间的抗压实验数据,采用三次多项式曲线模型对影响混凝土抗压强度值的变化龄期进行拟合分析时,由于法方程系数矩阵可能存在严重病态,从而导致拟合结果不可信。针对这种现象,应用2种不同岭参数k(岭迹法和双h公式法)的岭估计拟合试验,相比于普通最小二乘法,基于双h公式法的岭估计能够更加减小数据误差对拟合结果的影响,从而达到与实际结果更吻合的效果,并能较准确地预测出不同时期混凝土的抗压强度值。
Through the experimental data of the silicate concrete compressive test during the 28-day period, when cubic polynomial curve model was used to fit and analyze the changing day of concrete compressive strength, the fitting matrix of the normal equation might be seriously ill-coefficient, which leads to the issue that the fitting result cannot be trusted. In response to this phenomenon, the ridge estimation fitting experiment using two different ridge parameters k(ridge and double h formula) was applied. Compared with the ordinary least squares method, the influence of data error on the fitting result can be reduced by the ridge estimation based on the double h formula, so as to achieve amore consistent with the actual effect and a more accurate prediction on the compressive strength value of concrete in different periods.
Through the experimental data of the silicate concrete compressive test during the 28-day period, when cubic polynomial curve model was used to fit and analyze the changing day of concrete compressive strength, the fitting matrix of the normal equation might be seriously ill-coefficient, which leads to the issue that the fitting result cannot be trusted. In response to this phenomenon, the ridge estimation fitting experiment using two different ridge parameters k(ridge and double h formula) was applied. Compared with the ordinary least squares method, the influence of data error on the fitting result can be reduced by the ridge estimation based on the double h formula, so as to achieve amore consistent with the actual effect and a more accurate prediction on the compressive strength value of concrete in different periods.
引文
[1] 王迎飞,黄雁飞,周丽美.高强混凝土力学性能研究及其相关性分析[J].武汉理工大学学报,2006,28(3):67-71.
[2] 杜勇刚,刘海峰,马荷姣,等.沙漠砂混凝土抗压和抗折强度试验研究[J].混凝土,2018,40(01):102-104+108.
[3] 金钟.运用最小二乘法原理建立经验公式在推算混凝土强度中的应用[J].江苏建筑,2010,30(05):100-101.
[4] 陈正宇,徐君民.基于岭估计的GPS水准多项式拟合方法研究[J].工程勘察,2014,42(10):66-68.
[5] 邱卫宁,姚宜宾,陶宝藻.测量数据处理理论与方法[M].北京:测绘出版社,2006.
[6] 吴杰,李明峰,余腾.测量数据处理中病态矩阵和正则化方法[J].大地测量与地球动力学,2010,30(4):102-104.
[7] 来园莉.关于岭估计的若干问题研究[D].武汉,武汉科技大学,2010.
[8] 陈晓停,曹兰杰,汪金花.岭估计在多项式曲面拟合GNSS高程中的应用[J].河北联合大学学报(自然科学版),2016,38(4):1-6.
[9] 王乐洋,许才军,鲁铁定.病态加权总体最小二乘平差的岭估计解法[J].武汉大学学报(信息科学版),2010,35 (11):135-140.
[2] 杜勇刚,刘海峰,马荷姣,等.沙漠砂混凝土抗压和抗折强度试验研究[J].混凝土,2018,40(01):102-104+108.
[3] 金钟.运用最小二乘法原理建立经验公式在推算混凝土强度中的应用[J].江苏建筑,2010,30(05):100-101.
[4] 陈正宇,徐君民.基于岭估计的GPS水准多项式拟合方法研究[J].工程勘察,2014,42(10):66-68.
[5] 邱卫宁,姚宜宾,陶宝藻.测量数据处理理论与方法[M].北京:测绘出版社,2006.
[6] 吴杰,李明峰,余腾.测量数据处理中病态矩阵和正则化方法[J].大地测量与地球动力学,2010,30(4):102-104.
[7] 来园莉.关于岭估计的若干问题研究[D].武汉,武汉科技大学,2010.
[8] 陈晓停,曹兰杰,汪金花.岭估计在多项式曲面拟合GNSS高程中的应用[J].河北联合大学学报(自然科学版),2016,38(4):1-6.
[9] 王乐洋,许才军,鲁铁定.病态加权总体最小二乘平差的岭估计解法[J].武汉大学学报(信息科学版),2010,35 (11):135-140.