Research On Crack Detection Technology Of Bridge Pier Based On ROV Robot

Underwater pier structures are prone to cracking due to the increase in service life and erosion by water flow.The most serious is producing different kinds of cracks,causing huge safety risks to the bridge.Rovs-operated remotely operated vehicles(ROVs)have become a highly reliable tool that replaces a range of operations operated by divers,including inspection of structures,detection of shipwrecks on the surface,and research of life on the surface.The complex underwater environment requires the ROV to have a good motion control system and attitude adjustment strategy.The scattering of light in turbid water leads to blur,distortion,and brightness reduction in underwater detection images.It is difficult to identify and classify the characteristics of bridge cracks,which further influences the relevant departments to formulate maintenance measures for underwater pier structures.To solve the above problems,this thesis carried out ROV design for detecting crack structure of underwater piers,research on ROV attitude control strategy,and research on crack image enhancement and feature recognition and classification algorithms for underwater piers.Specific work contents are as follows:(1)Pier crack detection ROVs were designed to meet the requirement of pier crack detection.The modular idea was used to design the ROV structure,and the finite element analysis software was used to check the strength of the sealed cabin and the frame.The results showed that the designed structure and strength met the requirements of underwater detection.According to the actual work requirements,the design and selection of other key components,and finally the completion of the virtual prototype assembly and physical construction.(2)Aiming at the problem of water flow interfering with ROV movement,an attitude control strategy is studied.STARCCM+ software was used to simulate the damping values of ROV in the states of swing,heave,and yaw motion,and the hydrodynamic parameters were obtained by fitting.The unsteady motion technology of the software is used to calculate the additional mass of the ROV,and then the specific motion control model and the three-degree-of-freedom dynamics model are derived.The fuzzy PID controller is simulated and compared with the traditional PID controller by MATLAB software.The results show that the fuzzy PID controller designed in this thesis not only has a faster response speed but also has better anti-interference,which ensures better stability of the ROV in water.(3)Aiming at the problem of poor quality underwater pier crack images,underwater image enhancement and feature recognition technology are studied.Firstly,the image fusion enhancement method is used to process the pier crack image in multiple steps,and weighted fusion is carried out.Secondly,from the subjective and objective aspects of evaluation,the results show that the image quality has greatly improved after processing.Thirdly,the data set of pier crack features is extracted by threshold segmentation and noise removal.Then,the Improved Grasshopper Optimization Algorithm(IGOA)was used to optimize the Support Vector Machine(SVM)for the four classifications of pier cracks.The results show that IGOA-SVM is used to identify and classify pier cracks with higher accuracy.(4)Carry out pool experiments on an ROV.Firstly,the whole prototype of the ROV is assembled.Secondly,joint debugging of related components.Finally,performance debugging and functional experiments are carried out in the pool.The results show that the functions of motion and observation meet the design requirements,and the effectiveness of the designed fuzzy PID controller for attitude control is further verified.