沥青路面结构信息监测数据挖掘研究

发布时间：2018-02-10 10:16

本文关键词： 沥青路面结构信息监测数据挖掘性能预估　出处：《哈尔滨工业大学》2015年硕士论文　论文类型：学位论文

【摘要】：对沥青路面结构的动力响应进行实测研究是澄清复杂环境下沥青路面损坏机理的重要手段。传感器监测技术和计算机技术快速发展,对沥青路面结构进行长期监测并进行大量数据积累成为可能。许多监测项目积累了大量的路面结构动力响应数据和环境信息数据。然而,现有的研究侧重于对监测数据进行规律性描述,未能有效分析利用海量数据。为此,本文把数据挖掘理论引入沥青路面结构信息监测领域,实现对路面结构信息数据的有效分析利用。研究海量数据处理方法,在数据挖掘的基础上分析荷载和复杂环境下沥青路面结构动力响应,进行沥青路面性能预估。主要研究内容和成果如下:首先,分析常用的沥青路面结构监测数据采集存储方法,研究有效数据判定、数据滤波、关键信息提取等海量数据处理方法。分析数据挖掘在道路工程中的运用及桥梁监测、隧道监测、边坡监测等相关领域的数据挖掘技术,引入沥青路面结构信息监测数据挖掘概念。研究适用于沥青路面监测数据的挖掘方法,对海量路面结构信息数据进行数据挖掘。其次,结合数值模拟方法挖掘温度数据,研究实际温度场下的动力响应。建立沥青路面结构温度场分析有限元模型,研究基于实测沥青路面关键位置温度数据的沥青路面结构温度场数值模拟方法。与实测温度场对比分析,验证基于实测温度信息模拟路面结构温度场的有效性。使用ABAQUS软件建立沥青路面结构动力分析三维有限元模型,赋予实际温度场计算其动力响应,研究实际温度场下沥青路面结构的动力响应分布规律,并与使用等温模型计算的动力响应对比,分析等温模型在沥青路面力学计算中的不足。基于项目前期研究经验、现场轮迹线位置调查及数值模拟方法,结合项目已埋传感器位置信息及实测响应大小,分析沥青路面结构信息监测关键位置。再次,采用时间序列法研究应变影响因素,根据动态称重系统实测车辆信息数据进行轴载谱分析。根据关联方法建立最大竖向应变与轴载的关系,把设计年限内的交通量按轴载分级。考虑温度、交通量、轴重等参数,研究基于实测残余竖向应变的车辙预估方法。设计室内车辙试验,制作三层复合车辙板并埋入光纤光栅传感器,实测荷载作用下的竖向残余应变和一定荷载作用次数下的车辙变形,验证基于残余竖向应变的车辙预估模型。最后,分析实测纵向应变、横向应变的统计分布规律,研究应变衰变规律。基于时间序列自回归模型进行响应衰变预测,研究基于响应衰变进行沥青路面性能衰变预估的方法。本文重点对海量沥青路面结构信息数据进行挖掘研究,采用数据挖掘方法研究实际温度场下的动力响应,分析响应影响因素,进行使用性能预估。提高了对监测信息的利用度。
[Abstract]:The research on dynamic response of asphalt pavement structure is an important means to clarify the damage mechanism of asphalt pavement in complex environment. The sensor monitoring technology and computer technology are developing rapidly. It is possible to carry out long-term monitoring and data accumulation of asphalt pavement structure. Many monitoring projects have accumulated a large amount of pavement structure dynamic response data and environmental information data. However, The existing research focuses on the regular description of the monitoring data and fails to analyze and utilize the massive data effectively. Therefore, this paper introduces the theory of data mining into the field of asphalt pavement structure information monitoring. Based on the data mining, the dynamic response of asphalt pavement structure under load and complex environment is analyzed. The main research contents and results are as follows: firstly, the common methods of collecting and storing asphalt pavement structure monitoring data are analyzed, and the effective data judgment and data filtering are studied. Analysis of the application of data mining in road engineering and the data mining technology of bridge monitoring, tunnel monitoring, slope monitoring and other related fields, This paper introduces the concept of asphalt pavement structure information monitoring data mining, studies the mining method suitable for asphalt pavement monitoring data, carries on the data mining to the massive pavement structure information data. Secondly, combines the numerical simulation method to mine the temperature data, The dynamic response of asphalt pavement structure under the actual temperature field is studied. The finite element model of temperature field analysis of asphalt pavement structure is established. The numerical simulation method of the temperature field of asphalt pavement structure based on the measured temperature data of the key position of asphalt pavement is studied. To verify the validity of simulating the pavement structure temperature field based on the measured temperature information, the three-dimensional finite element model of asphalt pavement structure dynamic analysis is established by using ABAQUS software, and the actual temperature field is given to calculate its dynamic response. The dynamic response distribution law of asphalt pavement structure under actual temperature field is studied, and compared with the dynamic response calculated by using isothermal model, the deficiency of isothermal model in asphalt pavement mechanics calculation is analyzed. Field wheel track location investigation and numerical simulation method, combined with the location information of buried sensors and the measured response size, analyzed the key position of asphalt pavement structure information monitoring. Thirdly, the influence factors of strain were studied by time series method. According to the measured vehicle information data of dynamic weighing system, the axial load spectrum is analyzed. The relationship between maximum vertical strain and axial load is established by the correlation method, and the traffic volume in the design life is classified according to axial load. The parameters such as temperature, traffic volume and axle load are considered. The rut prediction method based on the measured residual vertical strain is studied. In the design room rutting test, the three-layer composite rutting plate is made and the fiber Bragg grating sensor is embedded. The vertical residual strain under measured load and the rut deformation under a certain number of loads verify the rut prediction model based on residual vertical strain. Finally, the statistical distribution law of measured longitudinal strain and transverse strain is analyzed. The response decay prediction based on time series autoregressive model and the method of predicting asphalt pavement performance decay based on response decay are studied in this paper. The dynamic response under the actual temperature field is studied by using data mining method. The factors affecting the response are analyzed and the performance prediction is carried out. The utilization of monitoring information is improved.
【学位授予单位】：哈尔滨工业大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：U416.217

【参考文献】