论文部分内容阅读
在充分调研国内外路面平整度检测技术的基础上,针对平整度数据检测中存在的异常数据、漏检数据等问题,结合其技术指标,基于最小二乘法的原理对试验数据的IRI测量重复性误差和IRI测量误差进行分析,确保数据源的准确性和可参考性,并对BP神经网络算法的有效性和实用性进行了验证,再采用BP神经网络方法对路面平整度检测数据进行识别和补充处理,使检测到的数据能反映被检测实体的真实信息。最后,结合工程实例,在Matlab上实现仿真和计算,把补充的数据和真实数据进行相对误差比较分析,结果表明:数据处理结果能很好地满足工程质量检测数据误差的要求。
Based on the full investigation of the pavement flatness testing technology both at home and abroad, aiming at the problems of abnormal data and missing test data in flatness data detection, combined with its technical indicators, the IRI measurement repeatability of test data based on the principle of least squares Error and IRI measurement error to ensure the accuracy and reference of the data source. The validity and practicability of the BP neural network algorithm are verified, and then the BP neural network method is used to identify the pavement roughness data. Supplementary processing, the detected data can reflect the real information of the detected entity. Finally, with the engineering example, the simulation and calculation are carried out in Matlab, and the relative errors of the supplementary data and the real data are compared and analyzed. The results show that the data processing results can well meet the requirements of engineering quality inspection data error.