Development Of Servo Control System For Rail Flaw Detection Based On Electromagnetic Tomography

Rail flaw detection is to detect the damage of rail by non-destructive detection technology.Railway accidents caused by broken rail often occur.In view of the safety of railway operation,rail flaw detection is significant.Currently,the mainstream front-end detection technologies are electromagnetic flaw detection and ultrasonic flaw detection.Hand push type ultrasonic flaw detector is used as the main method,and the large rail flaw detection vehicle as a supplement in domestic flaw detection task.The ultrasonic flaw detection method focuses the flaw in inner rail and rail waist.However,the surface of the rail and the bottom of the track belong to the blind area for ultrasonic flaw detection.The detection accuracy is high;the missing rate is very low for hand push type ultrasonic flaw detector.But it needs to clean the rail surface in the process of detecting,and ensure that the probe keeps in close contact with the rail surface.Thus,it needs to be operated by experienced professionals.So the detection efficiency is very low,and harsh environment is need for rail detection.The existing flaw detection equipment can not meet the demand of safety for railway operation.Therefore,a new type of onboard rail flaw detection scheme is designed and implemented in this thesis,and verified in Chawu Permanent Way Depot,which is significant in practice.The main work of this thesis is as follows:(1)A new kind of onboard rail flaw detection scheme whose electromagnetic probe is miniaturized and on-boarded is designed.With the train running,the distance of 5mm between the electromagnetic probe and the rail can be kept in real time to eliminate the effect of vehicle vibration to the measured signal.Thus,the on-line real-time flaw detection is implemented.(2)Mechanical part:the platform with three adjustable degrees of freedom consists of many mechanical components surrounding the two axes and electromagnetic probe.Innovations include:the design of L-type probe,the design of vertical anticollision ski,the design of rack by horizontal fishing method.L-type probe can make the coil cover the front and side of the rail at the same time.Vertical anticollision ski is used to avoid the probe over the track when the two rail joints with the collision.Horizontal fishing method to ensure that the probe chassis with vibration force adaptive through the fork.(3)Hardware part:the control circuit box including linear power supply circuit,voltage conversion circuit,motor drive circuit,communication interface circuit,speed sensor,eddy current displacement sensor is designed and built.(4)Software part:first of all,transplant Linux operating system to iMX283 controller;write part of the hardware drivers.Next,the communication interface,data analysis storage,eddy current measurement,closed-loop control system,control algorithm,multiple threads and other code is implemented.Finally,the Qt GUI is designed to show the motion state of the sensing platform and set the control algorithm's parameters.(5)Control algorithm:analysis and calculation of vehicle body vibration,the equivalent space force model of probe is analyzed.The PID control algorithm whose parameters are actually debugged when the train is running at low speed is impmented in the controller of iMX283.This thesis explores the iterative learning PID control algorithm and debugs the algorithm parameters on the microcontroller.Finally,the completed onboard flaw detection sensor platform is successfully installed on the low speed train,and the servo control is realized.The main purpose is to verify the feasibility of the flaw detection scheme,and lay the foundation for further application in high speed train.