| Catenary undertakes the important task of supplying energy to electric locomotives as an important part of electrified railway,and its running state directly determines the safety and benefit of electrified railway operation.The environment is harsh in some areas,and the catenary will produce a galloping phenomenon with low frequency and large amplitude under its influence.The galloping of catenary will not only affect the normal flow taking of trains,but also pose a serious threat to the safety of catenary,ranging from short circuit and trip to damage supporting devices and insulators,resulting in opening and disconnection of contact wires,and even serious safety accidents such as pole reversal.This thesis studies the non-invasive monitoring method of catenary galloping to prevent the occurrence of secondary disasters caused by catenary galloping.Video is used to realize on-line monitoring of catenary status in a non-invasive way to reduce the risk brought by galloping monitoring device to electrified railway operation.The image coordinates of the target point to be measured are extracted by video image technology,and the vibration amplitude of the contact line is obtained after coordinate transformation and calculation.An alarm signal is sent to start emergency processing after judging the galloping.Specifically,it includes the following methods:Firstly,the system scheme of non-invasive monitoring of catenary galloping is designed.The galloping monitoring method based on image analysis is determined by comparing the advantages and disadvantages of three monitoring methods of conductor galloping,combined with the structural characteristics and working requirements of the catenary.The coordinate transformation formula of a certain point on the contact network is derived according to the installation position information of monitoring equipment and camera imaging model,which lays a foundation for the subsequent extraction of wire galloping amplitude.Secondly,the extraction method of contact wire galloping amplitude is studied.The filtering method of the monitoring image is studied and experimented aiming at the image noise generated in the transmission process of the monitoring image.The median filtering method is selected for the monitoring image preprocessing stage according to the PSNR value and subjective effect evaluation index.Three methods of extracting the amplitude of contact wire galloping based on straight line detection method,inter-frame difference method,and template matching method are studied,taking it as the monitoring parameter of contact wire galloping considering that the amplitude of contact wire galloping is easy to measure and intuitive.The amplitude extraction effects of the three algorithms are compared by using the simulated conductor galloping experiment in the laboratory environment.The inter-frame difference method is selected as the extraction algorithm of contact wire galloping amplitude in the monitoring system according to the experimental results.Finally,a video monitoring system for catenary galloping is designed and developed by using the Tkinter framework and tested by using a simulated conductor galloping data set.The experimental results show that the on-line monitoring of catenary galloping can accurately extract the vibration information of the contact line,and can send out early warning information according to the set threshold on the premise of not affecting the safe operation of catenary,to carry out targeted processing when the catenary galloping occurs,which is beneficial to reduce the harm caused by the catenary galloping and provides a practical and effective processing method for the catenary galloping processing. |
