基于车载检测技术的水泥路面多级断板预防对策研究

Aiming at the problem of preventing slab broken of the cement concrete pavement, taking vehicle detection techniques as the breakthrough point, a research series are carried out, including the recognition of joint sealant damage based on the neural network, the detection of faulting based on the characteristics of relative altitude curves in the vertical section, the extraction of low-contrast multi-cracking under complex noise and the research on disengaging detection technology based on the GPR (Ground Penetrating Radar) image analysis, to build a high speed disease capturing, automatic processing and historical data management system based on a slab unit. By analyzing the micro and macro of all kinds of diseases in the observed sections, the time development rules of disease, the evolution regularity of the spatial distribution and the mutual influence will be revealed. On this basis, the disease forecast models are set up to identify the critical states of broken, utilizing the slab void as the core, the damage of sealant as the origin, the faulting as the significant complement and the multi-cracking as the interception line. Taking this as the breakthrough point of maintenance work, detailed daily maintenance countermeasure based on the vehicle fast test techniques and the model of slab crack warning and preventive maintenance countermeasure, which treating slab void to alleviate the adverse stress state of slab, using fillers curing to improve the distribution of slab void and using virtual IRI technique to relieve the dynamic loading caused by dislocation are proposed. In this way, a multi-prevention system is established to delay the slab broken, which has great significance to promote the maintenance level of concrete pavement, to extent road service life, to reduce maintenance cost and to make maintenance specification revision.

针对水泥路面断板预防的难题，以车载检测技术为突破点，开展基于神经网络的填缝料损坏识别、基于纵断面相对高程曲线特征的错台检测、复杂噪声下低对比度细微裂缝提取及基于探地雷达图像分析的脱空探测技术研究，构建板块为单元的病害快速采集、自动处理、历史数据管理系统；通过对观测路段各类病害微观、宏观分析，把握病害时间发展规律、空间分布状态的演变规律及其相互影响，在此基础上，建立病害预测模型，确立以脱空为核心、填缝料损坏为源头、错台为重要补充、细微裂缝为拦截线的断板预防性养护时机；并以此为养护作业的切入点，提出以脱空治理消减板块不利受力状态、以填缝料养护改善脱空分布、以虚拟IRI技术缓解错台引起的动载的预防性养护对策及依托车载快速检测技术和板块裂缝预警模型的精细化日常养护对策，建立多级断板预防体系，达到迟滞断板的目的。对提高我国水泥路面管养水平、延长寿命、降低养护成本及修订养护规范具有重要意义。

水泥路面是我国主要的路面形式之一，而断板是水泥路面结构破坏的主要形式，难以达到设计寿命，实践表明，采用预防性养护策略能够大幅度的迟滞断板的到来，能够有效降低全寿命周期养护成本，提高水泥路面寿命。通过项目研究，提出了基于数字图像和神经网络的填缝料识别及分类方法，基于激光断面仪纵断面曲线频谱特征的错台识别方法，高噪声、弱对比度下水泥路面裂缝及噪声剔除方法，基于深度学习的脱空探地雷达图像识别方法，并开发了接缝板角错动量检测装置，实现脱空板块的探测，构建了板块为单元的病害快速采集、自动处理、历史数据管理系统，实现检测速度快、自动化程度高、准确度好、细节和空间描述能力强的水泥路面断板及肇因的成套检测及分析技术。依托车载快速检测技术，在准确度高、历史脚印完整、影响要素齐全的分析数据基础上，采用微观分析与宏观分析相结合、时间要素与空间要素相结合的分析手段，充分把握各类病害时间发展规律、空间分布状态演化规律及其相互影响关系，建立了裂缝、填缝料、错台和脱空的路段病害预测模型和裂缝板块预警模型，在此基础上，提出了以脱空为核心，填缝料损坏为源头、错台为要素，细微裂缝为拦截线的断板预防性养护时机确定方法，并以脱空治理改善板块受力状态、以填缝料养护改善脱空分布状态、以虚拟IRI技术消减错台引起的动载的预防性养护对策及依托车载快速检测技术和板块裂缝预警模型的精细化日常养护对策，形成日常养护与预防性养护相协调、快速检测与基层执行力和宏观决策相统一的多级断板预防体系。项目成果可实现水泥路面裂缝及其肇因的快速检测，并迅速寻找最佳断板养护时机，从而有针对性地选择水泥路面断板预养护技术，起到防患于未然，降低全寿命周期成本；还可缓解我国优质沥青资源稀缺局面；除为公路养护管理部门检测评定路面状况和制定修复对策提供方法之外，还可以为修订或增补水泥混凝土路面养护技术规范提供最新依据。