Research On The Equivalent Conicity Of The Wheel Based On Computer Vision

The relationship between wheel and rail,as a link in the vehicle track system,determines the running characteristics of the railway vehicles.In wheel rail relationship,equivalent conicity is the most capable parameter to express the characteristics of wheel rail contact interaction.It has become an important parameter to measure vehicle’s stationarity.In various factors,Wheel tread curve change is the main reason leading to equivalent conicity variation.With the improvement of heavy haul and high-speed technology,as the carrier and driving force of vehicle weight and the transmission of braking force,the wear of wheelset is increasing.When the detection and lathing work are more frequently,fast and accurate measurement and equivalent conicity calculation for the repaired wheel pair are required.Determine whether the repaired wheel is in accordance with the standard to ensure the stability and safety of railway transportation.With the rapid development of railway technology,accurate and efficient wheel state measurement technology has become an important guarantee for the safe and stable operation of the train.The requirements for the precision and efficiency of wheel technical state measurement are also increasing in the fields of wheel rail relationship scientific research and train safety monitoring and so on.Therefore,it is of great significance to study the more fast and accurate automatic test and calculation system of wheelset.In order to calculate the wheel and the equivalent conicity quickly and accurately,a fast and efficient and fully automatic camera calibration method is proposed on the basis of the existing camera theory model and calibration method in this paper.A set of automatic wheel detection and calculation system which is based on computer vision is built in this paper.And this system is used for the detection and calculation of the repaired wheels.In this thesis,the HALCON program language is used to deal with the digital image of the wheel tread images collected by the computer vision system,and the complete and smooth wheel tread curves are obtained.Then the camera calibration parameters are used to convert the pixel coordinates of the wheel tread curve to the world coordinates,and to obtain the contour coordinates of the wheel tread.On the basis of the calculation method of the equivalent conicity of UIC519,an improved calculation method is put forward.In this paper,the Simpack and Microsoft Visual Studio software are used to calculate the wheel diameterΔr,the snake shape sinusoidal wave lengthλand the equivalent conicity tanγ_e.Finally,the standard data is used to verify the detection and calculation methods of this paper.The results show that the accuracy of the detection and calculation method proposed in this paper meets the requirements,which can provide the basis for the parameter measurement and the effect evaluation of the repaired wheel.