基于检养系统的病害原因和养护对策智能诊断技术研究

Research on Intelligent Diagnosis Technology of Defects Causes and Maintenance Countermeasures Based on Inspection and Maintenance System

为解决现有桥梁病害成因和养护对策严重依赖专业技术人员经验判断的问题,同时提高公路桥梁检测养护领域的智能化和自动化水平,以现有的研究内容及成果为基础,基于人工智能、机器学习等新一代信息技术和检测、评定、养护等专业核心技术,通过构建病害基本信息库、病害成因分析库、病害养护处治方案库,并结合特征词提取技术、模糊概率匹配算法和匹配准确度提高算法,实现对公路桥梁外观检测病害的成因和养护对策的智能匹配。目前已完成108种梁桥病害、210种病害成因、189种养护对策和1 033个特征词的构建,通过不断测试、积累和优化计算,使梁桥的病害成因和养护对策匹配准确率已达85%。将该研究成果应用于桥梁检养专业系统,以某混凝土空心板梁桥的板底横向裂缝为例进行病害结果、成因和养护处治对策诊断,其诊断结果可满足实际工程需求。研究成果可快速部署并应用于公路桥梁检养专业系统,实现桥梁病害原因和养护对策的快速诊断,实现桥梁检养的系统化、智能化和快速化,同时可借鉴此研究方法拓展其他桥型和其他专业的智能诊断技术研究,具有很好的沿用性和借鉴性。

In order to solve the problem that the causes of the existing bridge defects and maintenance countermeasures mainly rely on the experience judgment of professional and technical personnel,and at the same time to improve the intelligence and automation in the field of highway bridge detection and maintenance,based on the existing research content and results,artificial intelligence,machine learning and other new generation of information technology and detection,evaluation,maintenance and other professional core technologies,by constructing the database of basic defects information,analysis of causes,maintenance and treatment schemes,and combining feature word extraction technology,fuzzy probability matching algorithm and matching accuracy improvement algorithm,the intelligent matching of causes of defects detection and maintenance countermeasures of highway bridges is realized.At present,108 kinds of girder bridge defects,210 kinds of defect causes,189 kinds of maintenance countermeasures and 1 033 features have been constructed.Through continuous testing,accumulation and optimization calculation,the matching accuracy of the causes of girder bridge defects and maintenance countermeasures has reached 85%.The research results are applied to the professional bridge maintenance system,and the results,causes and curing measures of the transverse cracks in the slab bottom of a concrete hollow girder bridge are diagnosed.The diagnosis results can meet the actual engineering needs.This research results can be rapidly deployed and applied to highway bridge inspection system,realize the rapid diagnosis of bridge defects causes and maintenance measures,and realize the systematization,intellectualization and rapidity of bridge maintenance.At the same time,this research method can be used for reference to expand the research on intelligent diagnosis technology of other bridge types and other specialties with good practice prospect and reference.