基于Bootstrap法的连续压实指标目标值计算方法

Method for calculating target value of continuous compaction index based on Bootstrap method

有效控制填筑体压实质量是保障道路、大坝、机场等基础设施使用寿命和安全运营的关键。随着连续压实控制技术的广泛应用,科学合理地确定连续压实指标目标值对于填筑体质量控制具有重要意义。提出了一种基于Bootstrap法的连续压实指标目标值计算方法,与基于有限个试验数据采用经验公式获取压实指标目标值的传统方法相比,该方法从统计学角度,根据常规质量验收指标与连续压实指标的经验公式建立了置信度为(1-α)×100%的回归模型置信带,并通过选取置信上限来确定连续压实指标目标值。最后,结合国内外文献中相关地基压实工程中的现场试验数据对该方法的实用性进行验证,结果表明:该方法计算得到的连续压实指标目标值要优于常规经验公式计算值,且具有更高的可信度。该方法能够更合理地对由小样本不确定性导致的压实质量风险进行补偿,且补偿值随试验数据相关性的减小而增加,从而使压实质量控制标准更加科学与可靠。此外,相比于已有文献给出的置信区间,该方法所得置信区间宽度更窄,避免了过盈估计,具有较好的工程指导意义和应用前景。

The effective control of compaction quality of the filling body is essential to ensure the service life and operation safety of infrastructures such as roads, dams and airports. With the wide application of the continuous compaction control technology, it is important to determine the reasonable target value of continuous compaction index to guarantee the compaction quality. A novel method for calculating the target value of compaction index based on the Bootstrap method is proposed. Compared with the traditional methods based on the empirical formula along with a limited number of test data, the proposed method statistically establishes a regression model confidence band with the confidence level of(1-α)100% based on the empirical formula, and then determines the target value of compaction index by selecting the upper confidence limit. Finally,the proposed method is verified through the existing field test data of foundation compaction projects from literatures. The results show that the target value calculated by the proposed method has a higher reliability compared to that by the empirical formula. Besides, the errors of compaction index caused by the uncertainty due to the limitation of samples can be amended by the proposed method, and the amendment value increases as the correlation of the test data decrease, which ensures the reliability of the control standard of compaction quality. In addition, compared with that of the confidence interval given in the literatures, the width of the confidence interval obtained by the proposed method is narrower, hence the over-estimation can be avoided, indicating that the proposed method is of significance in both the current engineering design and the prospective application.