Research For Hunting Instability Detection Of Bogie By Lateral Motion Identification Method

Hunting is one of the natural characteristices of railway vehicles.Its main reason for hunting is the taper of wheel tread.Due to the turning requirement,it's impossible to avoid the hunting motion.However,railway vehicle easily comes into an instability state,even derailment when its running speed is over critical speed.Therefor,hunting monitored equipment is essential for ensuring the running safety.So far,the general route to monitor the hunting motion is on-board monitored system developed based on the standards,for example EN14363,UIC518 and so on.In these standards,the lateral accelerations of frame are regarded as the input,a threshold is set to classify stability and instability.However,these methods have already shown some shortcomings.Firstly,due to the setting of acceleration threshold,these methods can only recognize the hunting instability when the amplitude of vibration gets highly enough.This means that when the method recognize the hunting instability,the bogie has already in extremely dangerous condition.Secondly,for supercritical scenario,if the vehicle runs at a speed between nonlinear speed and linear critical speed,it may make the bogie run with low amplitude motion.The amplitude of vibration may much lower than the threshold for alarm.In practice,the vibration is adverse for the safety of railway vehicle.But these hunting-monitored method can't recognize it.Thirdly,the critical point for derailment of bogie is the lateral motion of wheelset.Under hunting instability condition,the peak of lateral motion for wheelset is commonly decided by amplitude of lateral acceleration and frequency of lateral vibration.However,the traditional methods haven't taken vibration frequency into consideration.Finally,frame is treated as rigid in the traditional methods based on frame acceleration.So,there is no requirement about the placed position of acceleration sensors.But the test result of EUM train shows that the different position of frame have different vibrations.Therefore,how to set sensor becomes a problem to deeply investigate.To solve the above-mentioned problems,a monitored method for hunting inatability based on the lateral motion of wheelset is proposed.The vibration signals of axle box are analyzed by the singular value decomposition with Hankel matrix.The peak value of wheelset's lateral motion can be calculated by main singular values.Those real test data show the new method can not only recognize the instability which traditional method can find,but also the low amplitude hunting which be treated as omen of instability.In term of the hardware of monitoring system,its core is the normal work of acceleration sensor.However,acceleration sensor is a typical small signal equipment which has special requirement for power supply quality.The normal power line filters can make most of on-board system meet the requirement of standard GB/T24338.However,for a vibration sensor which only has millivolt level output,the normal power line filters lack ability to keep vibration sensor working well.For the system mentioned with the normal power line filter equipped in the article,the output error reaches to 1.5g under the EFT test with GB/T24338's requirement.To find the cause of error rising,the real test results and simulation results are compared.A new filter based on transformer is designed,instead of the traditional power line filters based on resistor and capacitance.With the new filter equipped,the system error reduces from 1.5g to 0.01g during EFT test with same jamming level.In addition,the new filter can also deal with radiation emission and conduction emission problems.