| By the end of 2020,the number of highway bridges in China has reached 912,800.A large number of bridges built in the early stage have entered a period of rapid performance decline,leading to frequent bridge accidents.With the further development of China’s economy,it can be predicted that China will be in the stage of massive bridge construction and management for a long period of time.At the same time,the bridge construction and management of such a huge volume will face huge technical and economic challenges.Therefore,the development of information technology and intelligent technology to promote bridge construction,management efficiency and level has become a general trend,and has been included in the development planning outline of the Ministry of Transport.Bridge monitoring,detection and performance evaluation are the main technical means to find bridge diseases in time and make correct treatment decisions.The timeliness and reliability of detection and evaluation are related to people’s life and property safety.However,the current detection and evaluation methods have strong manual dependence,insufficient reliability,and low degree of informatization and intelligence.There are still many industrial problems that have not been solved theoretically and technically,which are difficult to meet the requirements of future mass bridge supervision in terms of safety,efficiency and cost.In view of the current pain points of bridge inspection and evaluation,and combined with the development trend of the industry in the future,this paper introduces the concept of digital twins,and studies the use of holographic scanning,photogrammetry and other digital reality capture technology to build the twin model consistent with the real bridge.Based on the new digital twin data type and digital information technology,a number of new bridge detection and evaluation theories and methods are developed.The main research contents and achievements of this paper are as follows:(1)The general definition,characteristics,data characteristics and reasonable basic data model of bridge digital twin are put forward,and the key technology research of bridge digital twin basic data model is carried out.The precise registration method and algorithm of bridge point cloud are developed,which can make the registration accuracy of point cloud reach sub-millimeter level.In the point cloud environment,an automatic parameter extraction algorithm for typical bridge components is developed to solve the problem of parameterized driving modeling of real bridges.Based on photogrammetry theory,a direct mapping algorithm of apparent damage from 2D image to 3D model is developed,which solves the precise modeling problem of bridge damage such as cracks.(2)The inversion method of twin model of bridge mechanics is proposed.The high fidelity geometric twin model and apparent damage twin model were used as input sources for force analysis,and accurate spatial continuous geometry and spatial continuous deformation model were used as constraint conditions to modify and iterate invert the force analysis model.A spatial continuous deformation mode calculation method based on holographic scanning is developed.Experimental results show that the method has a real bridge deformation measurement accuracy of 0.1 mm.The reliability of the proposed inversion method for mechanical twin model is verified by designing a scaled concrete arch bridge.(3)An intelligent inspection method of component level fine apparent damage based on real bridge twin model is proposed.A reliable modeling strategy for 3D bridge UAV photogrammetry was developed,and the accurate construction of 3D twin geometric model of the bridge was realized by combining the terrestrial laser scanning point cloud.Aiming at typical bridge component types,a universal component-level fine inspection 3D route planning algorithm is developed.Taking DJI UAV as the carrier,the inspection task generation algorithm is developed.The reliability of the proposed inspection method is verified in a large overpass.(4)The detection and control method of segmenting bridge digital preassembly is proposed.Taking steel tube arch bridge as an example,an accurate reverse modeling parameter extraction algorithm for precast steel tube arch segments in point cloud environment was developed to realize the synergy between scanning model and 3D finite element model.An assembly simulation method which can accurately consider the redistribution of internal forces in the process of bolt forced assembly is developed to predict the mechanical state of prefabricated components in the process of virtual assembly.For prefabricated components containing manufacturing errors,an optimal correction calculation method for prefabricated components without stress parameters is proposed based on the idea of stress state method.For the arch forming process of steel tube arch rib with manufacturing errors,the optimization calculation model of cable force in the whole process is put forward,and the problem of one-step cable adjustment in the whole process is solved.The proposed digital preassembly detection and control method has been verified in a steel tube arch bridge with a span of 507 m.(5)The calculation theory and realization method of cable force based on twin model of stay cable is proposed.Based on the catenary equation,the nonlinear regression parameter estimation model of measured cable shape and cable force is established,and the calculation formula of cable force iterative method is given.Based on the holographic scanning point cloud data,the automatic segmentation and cable shape extraction algorithms of cable point cloud are developed.The real bridge experiment shows that compared with the frequency method,the efficiency of scanning cable measurement method is improved by 8 times.The cable error factors and their influence on the calculation accuracy of cable force are analyzed in detail.The results show that the cable length,boundary conditions,sectional stiffness and local noise are not affected by the cable shape method.(6)The theory and implementation method of reverse evaluation of existing suspension bridge stress state based on twin cable shape of main cable are proposed.The calculation theory of suspender force of suspension bridge based on measured main cable intersection is established,and the correctness of the theory is verified by experiments on two real bridges.An automatic algorithm is developed to calculate the coordinates of the main cable intersection in the scanning point cloud.An inversion method of stress state of suspension bridge based on influence matrix method is proposed on the basis of main cable twin cable.An indoor experiment and a real bridge experiment verify the reliability of the inversion method. |
