| During the operation of long-span girder bridges,certain problems such as cracking and long-term deflection may exert significant influences on the safety and application prospect of such bridges.For one thing,it is of urgent necessity to ponder the causes of these issues from the perspectives of design,construction,and maintenance.For another,a monitoring system is also required to track and evaluate the structural performance to grasp the structural status and prevent sudden structural damage and collapse.Deflection is the most intuitive index reflecting bridge status,which therefore plays a vital role in monitoring and analyzing the dynamic deflection of the bridge under the action of operational loads.This dissertation probes into connected pipe based dynamic deflection monitoring technologies,dynamic deflection separation algorithm,deflection based moving load identification method,as well as deflection informed structural status analytical method.These methods are applied to a long-span prestressed concrete continuous girder bridge in an attempt to form a deflection informed methodology of long-span girder bridges including monitoring technology,data analysis,load identification,as well as status evaluation.Main contents and findings are as follows:(1)The key technologies for bridge deflection monitoring based on a connected pipe system is studied.To begin with,a bridge deflection test model based on a connected pipe system is established,and its monitoring effects on structural static and dynamic deflection are analyzed;Then,the influence of bridge vibration on dynamic deflection monitoring with the help of connected pipe system is mathematically derived,the results of which are verified by model tests;In the end,the influence that the connected pipe system’s layout may impose on the dynamic deflection monitoring precision is investigated with several precision improvement measures proposed,and meaningful suggestions for the construction of connected pipe system for realistic bridge deflection monitoring obtained.According to the results,it can be concluded that the measurement results of the static and dynamic deflection in the model test fit well with these by a referred displacement gauge,which verifies the monitoring accuracy of the connected pipe system.Besides,structural vibration is influential to the pressure acquisition of the connected pipe system,especially to an oblique pipeline.It worths mentioning that such influence can be well eliminated by theoretical derivation,which has been verified by the model test.Moreover,the dynamic deflection of the connected pipe system is independent of the pipeline step but is significantly affected within 1.5m of the curved region.These characteristics should be taken into account in the authentic connecting pipe layout to ensure accurate monitoring of the structural deflection.(2)A sparse separation is proposed based on a joint dictionary algorithm of bridge dynamic deflection.In the first place,atoms are generated with the features of various types of deflection signals being considered,and a joint dictionary is constructed in combination with spare regularization mode.Various types of deflection signals are reconstructed via these dictionary atoms,and the algorithm is established to separate the time domain of mixed deflection signals.Afterward,what kind of separation effect the proposed algorithm will impose on dynamic deflection is studied,including vehicle loading effect,temperature effect,and long-term effect.By the way,the impact of noise on the separation effect is parametrically studied.The results demonstrate that the proposed joint dictionary dynamic deflection separation algorithm is equipped with favorable accuracy,effectiveness,and noise-resistance,which guarantees the effective separation of field monitored bridge dynamic deflection.Since the joint dictionary has taken various types of deflection signals into account,it is more flexible than a single dictionary in terms of application ability.(3)Following the dragonfly algorithm,an on-bridge moving load identification method is put forward.To begin with,the vehicle load and passage time along the bridge are deemed as two variables for identification,and the optimization problem is established by comparing the differences between the realistic deflection and calculated deflection of the bridge.Later on,the dragonfly algorithm is introduced to solve the optimization problem,which constructs the moving load identification method.Then,the proposed method is carefully verified by numerical examples of a simply-supported girder and a continuous girder,and the influence of pavement roughness and noise on the identification results is investigated as well.Finally,this method is applied to a realistic bridge which carries a moving test and a jump test by a heavy truck.Based on the results,it can be claimed that the on-bridge moving load identification method that is proposed based on the dragonfly algorithm possesses decent accuracy and efficiency,and can well capture the passage time and vehicle load simultaneously.Besides,this identification method is not influenced by the noise and pavement roughness,which is representative of robustness.As for the actual engineering application,the identification error of passage time and vehicle load is only 3%,indicating its prospect for engineering application.(4)Based on separated deflection and Dempster-Shafer evidence theory,a structural status analysis method is proposed.At the very beginning,on account of a cellular automaton,microscopic random traffic load modeling method is established which incorporates the intelligent driver model of car following,acceptance distance of lane-changing,equivalent dynamic wheel load method,and so on.The method is applicable to the analysis of the bridge deflection features under random traffic flow and bridge interaction analysis,which intends to establish the hierarchical early warning method of vehicle deflection component.Then,the hierarchical early warning methods of temperature deflection component and long-term deflection component are established through finite element analysis and literature review.At last,the Dempster-Shafer evidence theory is applied to the structural state analysis based on separated bridge deflection,to deal with the fusion and conflict of multi-source information and form a hierarchical evaluation system of bridge status.It is noteworthy that the proposed structural state analysis method can make full use of different separate components of the monitoring bridge deflections and take the advantages of the Dempster-Shafer evidence theory for the combination of multi-source information,which can be consequently further extended to the application of any bridge system.(5)The above methods are applied to an actual engineering project,which is concerned with a long-span prestressed concrete girder bridge equipped with a connected pipe based dynamic deflection monitoring system.In this case,the monitoring effect and accuracy of the connected pipe system for engineering application are analyzed.Besides,it also researches the effect and accuracy of the joint dictionary method in the separation of dynamic deflection for engineering application.Furthermore,the application procedures of the bridge status hierarchical evaluation method based on Dempster-Shafer evidence theory is refined accordingly.Results indicate that monitoring deflections with a connected pipeline are in good agreement with these using artificial leveling method for the project.Additionally,the deflection separation algorithm can well obtain the deflection components of vehicle effect,temperature effect,and long-term effect.Combined with the deflection monitoring data of the bridge in the past three months,the structural status of the bridge is analyzed,and is found to be in a good state. |
PDF Full Text Request