Construction Of Historic Timber Structure Benchmark Platform And Research On Monitoring Method

Cultural self-confidence is the most basic,broadest,and deepest self-confidence of the Chinese Nation,and it is also a necessary condition for realizing the Chinese Dream.The protection of cultural relics is the basic requirement of cultural self-confidence.The most critical part of the protection of cultural relics is the protection of historic timber structures(HTS).Under the guidance of scientific and technological protection,professionals with multi-disciplinary backgrounds have joined the preventive protection of HTS.Structural engineering professionals have protected HTS with multiple levels of damage from the perspectives of mechanism research and field work.However,in the process of field work,the structural health monitoring(SHM)method is still in its infancy.Although the health monitoring methods in the field of bridge structures,super high-rise structures,and large-span spatial structures have been relatively complete,HTS has an extremely long service period,extremely high values,extremely special excitation response,extremely complex damage forms,extremely complex structural form,extremely unique materials,and extremely strong uncertainty,these characteristics hinder the development of health monitoring methods in the field of HTS.Therefore,this article is guided by the evaluation and identification indicators to establish a health monitoring system of HTS,including displacement monitoring method,crack monitoring method and strain monitoring method based on computer vision technology and data mining technology.Establish the first benchmark platform for health monitoring of HTS as the verification platform of this system,so that the above system can serve the real HTS.(1)This dissertation establishes the first benchmark platform for health monitoring of HTS,the platform includes physical model,numerical model and supporting website.The basic composition and structural functions of HTS are retained,which eliminates the impact of HTS particularity on the development of health monitoring.In the process of establishing the numerical model,a constitutive relationship model based on phenomenological theory is proposed,which makes up for the current shortcomings.Then,a new lattice model which can accurately simulate the local heterogeneity and local damage of HTS is proposed.The above constitutive model and the improved lattice model are used to complete the construction of the numerical model.Finally,the dynamic test method is used to obtain the dynamic parameters of the platform,and verify the availability of the platform.(2)A displacement monitoring method for HTS has been established.The method fully considers the displacement monitoring requirements of HTS,and the method independently develops scripts based on the open-source motherboard.The first visual equipment for displacement monitoring of HTS is made.On the basis of hardware,the method optimizes the binocular stereo vision algorithm based on Zhang Zhengyou calibration method and feature point tracking algorithm.The method completes the accurate three-dimensional coordinate tracking of key points,and realizes the non-contact displacement monitoring of HTS.(3)A crack monitoring method for HTS has been established.The method fully considers the uniqueness of the crack monitoring process.Relying on the flight controller and open-source platform,a crack monitoring UAV system which can fly with high precision in low GPS environment is made.Based on the above hardware,a camera system including distortion correction algorithm,pixel resolution calibration algorithm,feature point extraction and matching algorithm are established.A graphics processing system including graphics preprocessing algorithm,new threshold segmentation algorithm and crack feature calculation algorithm are established.The method greatly reduces the error and cost of crack monitoring,greatly improves the accuracy and efficiency of crack monitoring,and realizes the contactless crack monitoring of HTS.(4)A strain monitoring method for HTS based on massive strain monitoring data is established.The method combines traditional machine learning technology and deep learning technology to identify distorted data.After decoupling the data,the prediction algorithm and extreme value theory are introduced to complete the short-term prediction of strain monitoring data,so as to improve the structural safety redundancy.Then,an analytical model is established,which integrates semi-rigid boundary conditions,orthotropic parameters and temperature sensitivity.The method has completed the unique shear stress monitoring of HTS,and realized the strain monitoring of HTS with high data mining rate.The above three monitoring methods constitute a complete HTS health monitoring system with evaluation and identification as an indicator.Through the mutual verification with HTS health monitoring benchmark platform,it fully proved the usability of the platform established in this dissertation,and also fully proved the effectiveness and accuracy of HTS health monitoring system established in this dissertation.Figures 95,Tables 43,References 266.