The Design And Implementation Of The Application Software For The Status Recognition Of Track Fastener

Rail transit operation safety has always been an important part of China’s traffic safety.Nowadays,the construction of high-speed railways and subways is developing rapidly.Rail transit safety operations have become an important manifestation of the stability and development of China’s rail transit,and also carry the wishes of millions of families.However,the current maintenance mode of rail transit is basically still in the traditional mode,such as visual inspection of rail fasteners by manual seeking.However,manual inspections in broad and dense track lines have problems such as low efficiency,strong subjectivity,and time-consuming and labor-intensive inspections.How to use modern technology in combination with the Internet and various sensors to efficiently and stably monitor all aspects of rail fasteners has become an important subject at home and abroad.This thesis designs and implements a fast spring bar status recognition application software system.The system uses a C / S and B / S architecture.The fastener recognition application software integrates a dual-line camera control module and displays the recognition results.Module,data submission module.Through the user-friendly UI interface,you can complete the functions of data collection,data storage,status recognition,and data submission for one-stop dual-line camera.In addition,users can log in to the track fastener data platform through the browser at any time,query the track fastener detection data,leave a message on the relevant data,and compare and analyze with the previous data.This set of application software system has strong comprehensiveness and practicality.The main work and results of the thesis are as follows:(1)Research on the state identification algorithm of fast spring bars of rail fasteners,and propose a fast spring bar positioning algorithm based on grayscale difference,which can locate fast spring bars of rail fasteners with good accuracy and speed.In addition,combined with the most popular convolutional neural network nowadays,this paper proposes a rapid detection method for rail fasteners based on computer vision SSD(Single-Shot Multi Box Detector).And use Python programming to achieve,generate exe executable file for the main program to call.(2)Use C # to develop line scan camera image acquisition and status recognition application software.The application software uses multi-thread technology,supports dual cameras to work at the same time,and can save image data in real time.In addition,the application software integrates orbital fast spring bar state recognition algorithm modules,and analyzes the collected image data to identify unfastened fasteners based on the texture features of the orbital fasteners.(3)Develop a web back-end microservice based on Spring Cloud to provide a data interface for the rail fastener data platform.Complete the construction of microservices,My SQL database development,design and development of various data interfaces,configure Nginx server to complete reverse proxy and proxy for static resource files.Ensure that the entire system platform can run normally.(4)Develop front-end module of rail fastener data platform based on Web frontend framework Vue.js,complete user login module and fastener data query module.Users can log in to the platform remotely,view the test data of different lines,and comment on the test results.