Structural Damage Detection Of Bridge Based On Improved Dragonfly Algorithm And Regularization

With the development of bridge engineering technology,many modern bridges with complex force systems and large spans have been put into use.However,with the increase of the bridge service time,along with the fatigue and aging of structural materials,the long-term effects of vehicle loads,natural environment erosion,wind loads,and earthquakes,the bridge's safety performance and durability will inevitably decrease.The structure will inevitably be damaged.Therefore,it is particularly important to strengthen the Structural Health Monitoring?SHM?of active bridges,and Structural Damage Detection?SDD?is the core content of bridge structural health monitoring.This paper combines the Dragonfly Algorithm?DA?and the L_1/2regularization method to conduct related research on bridge structure damage detection.The specific research content is as follows:?1?The research status of structural damage detection at home and abroad is expounded systematically.At the same time,the research status of dragonfly algorithm and regularization method are introduced in detail,and the main breakthrough points and innovation points of this paper are determined.?2?The principle of the dragonfly algorithm and the theoretical basis of damage detection are introduced in detail.The original objective function is constructed using the frequency residual and MAC.Based on the dragonfly algorithm and the original objective function,the damage detection numerical simulation is performed on the simply supported beam.The results show that the dragonfly algorithm can be used to identify the damage of the structure;However,for specific positions of the structure and multiple damage conditions,the dragonfly algorithm's ability to optimize is insufficient,at the same time,the original objective function has insufficient anti-noise capability,the recognition accuracy under noise needs to be improved.?3?Aiming at the problem of insufficient optimization ability of dragonfly algorithm and easy to fall into local optimum,this paper proposes a Nelder-Mead Dragonfly Algorithm?NMDA?.The NMDA is tested using 6 different standard test functions.The test results show that the NMDA have stronger optimization ability and can overcome the local optimization problem.In addition,the previous simple beam model using the improved algorithm was used to perform damage detection simulation again.The simulation results show that under the same working conditions,the improved algorithm has stronger damage detection ability,which verifies the superior performance of the NMDA.?4?Aiming at the problem that the anti-noise ability of the objective function based on frequency and mode is not strong,this paper constructs a new objective function based on the L_1/2 regularization method.Using the improved NMDA through numerical simulation of damage detection of simple supported beams,cantilever beams and continuous beams,the results show that the L_1/2 regularized damage detection method based on NMDA can effectively overcome the impact of noise and accurately identify the real damage.?5?In order to verify the applicability of the method proposed in this article,a simple supported beam model of aluminum alloy was set up in the laboratory.By simulating the damage of the simply supported beam,the five frequencies and vibrations of the simply supported beam were collected using equipment under each working condition.The collected data is imported into the NMDA for damage detection.The test results show that the NMDA of L_1/2 regularized damage detection method proposed in this paper can accurately identify the true damage of the test beam,which has good engineering practical value.