Dynamic Reference Measurement System For Intelligent Track Inspection Vehicle

With the large number of subway lines open,the safety status detection of the subway tunnel infrastructure is becoming more and more important.The track infrastructure intelligent inspection vehicle in this thesis can realize the automatic comprehensive detection of the safety state of subway tunnel deformation,crack and settlement.However,vibration occurs during the running of the inspection vehicle.The uncertainty of the vehicle attitude will greatly affect the measurement accuracy of the tunnel deformation detection system.The tunnel deformation measuring equipment is installed on the vehicle,but the deformation detection needs to use the track as the reference coordinate system.Therefore,it is first necessary to measure the vehicle attitude relative to the track,and then the vehicle body attitude compensation is performed on the measured tunnel profile data to reduce the error caused by the vibration of the vehicle body during tunnel deformation detection.In this thesis,a dynamic reference measurement system for the intelligent track infrastructure inspection vehicle is developed.Four two-dimensional laser scanning sensors are used to collect the profile of the rail section.The measurement track plane is calculated by extracting the four gauge points.By comparing the measured track plane and the standard track plane,the positional relationship between the vehicle body and the track reference is calculated,and then the vehicle instantaneous attitude is obtained.Finally,the tunnel profile data is compensated by the body vibration.This dynamic reference measurement system can greatly improve the measurement accuracy of the tunnel profile deformation.This thesis presents the measurement principle and implementation method of the dynamic reference measurement system for the intelligent infrastructure inspection track.The key research contents include proposing a measurement data coordinate system transformation method which combines the curve center fitting algorithm and the improved ICP algorithm.An adaptive median filter based on gray correlation is used to filter the measurement data by introducing gray correlation coefficients.The thesis uses a method of constrained least squares arc fitting to fit the top curve of the rail.After threshold judgment and precise segmentation,the quadratic fitting of the rail top curve is performed.Four gauge feature points are extracted and the measured track plane is obtained.By comparing the measured track plane with the standard track plane,the vehicle attitude is obtained.The tunnel profile data is finally compensated for vibration.In order to verify the effectiveness of the system,the dynamic reference measurement system of the inspection vehicle is tested in the subway tunnel model and the Xia Jiao metro station tunnel site of the Guangzhou Metro Line 3.The accuracy of gauge feature points extracting is verified by measuring the gauge.The gauge results calculated by system are compared with the track gauge measurement results,and the gauge measurement error is within±1mm.The experimental results show that the system can effectively the system can effectively compensate the vehicle vibration for the tunnel profile data.The system can improve the detection accuracy of tunnel deformation and meet the needs of field application.