Research On Transmission Model And Fault Diagnosis Method Of Track Circuit

The track circuit is a circuit with a rail as a conductor for automatically and continuously detecting whether the line is occupied by the train.In addition,the track signal transmitted by the track circuit also provides movement authorization for the train by electromagnetic induction.The role of the track circuit mainly includes:(a)Supervise the occupation/clearance of the track section.(b)Send train control information to the train.(c)Check the broken track fault of the track section.The type ZPW-2000 track circuit is widely adopted in the China highspeed railway.An increase in the propagation distance of a track signal is achieved by installing compensation capacitors between the rails and electrical isolation with the adjacent track circuit is implemented using electrically insulated joints.The working performance of track circuit largely dependents on the transmission quality of the track signal and it directly affects the transportation safety and efficiency.On the one hand,establishing a track circuit transmission model is an effective way to analyze the performance of the track circuit.The transmission model of track circuit is the theoretical basis for analyzing the transmission performance of track circuit.The research on the transmission model of track circuit plays an important role in mastering,understanding and analyzing the transmission characteristics of track circuit.For example,the key technical parameters of the track circuit influence each other and must be taken into consideration when determining the parameter value.The transmission model of the track circuit can calculate and analyze the key technical parameters of the track circuit under specific conditions,which will provide theoretical support for the design and improvement of the track circuit equipment.In addition,from the perspective of operation and maintenance,the track circuit transmission model is also of great significance,and the influence of the environment and the fault on the performance of the track circuit can be analyzed by means of the transmission model.The track circuit is an important basic safety equipment for the development of China railways.The failure of the track circuit can affect the transportation efficiency,which may have catastrophic consequences.Therefore,the safe operation state of the track circuit must be mastered at all times.At present,in order to better grasp the status of signal equipment the Railway Bureau has successively activated the signal centralized monitoring system,the inspection vehicle and the locomotive signal remote monitoring system.The monitoring system collects important operational parameters of the track circuit in real time through a specific data collection device.These data can provide support for device fault diagnosis.Although the existing monitoring system provides a method for judging the working state of the track circuit,unfortunately,these methods are all threshold-based methods,and do not dig deep into the potential relationship between the massive monitoring data and the state of the track circuit.Therefore,in practical applications,it can only be known whether the device deviates from the normal state and cannot accurately locate the fault.Based on the above analysis,it is necessary to study advanced fault diagnosis algorithms on this basis to further improve the intelligence level of existing monitoring systems.This dissertation aims to improve the theoretical analysis level and maintenance efficiency of the track circuit.The main work and conclusions are as follows:(1)The problem of transmission characteristics analysis of integrated track circuit(ITC)in the station was studied.The transmission model of ITC was established based on the theory of uniform transmission line.Each part of ITC was represented by equivalent two-port network.Based on the transmission model,the influence of outdoor environment(such as ballast resistance and train shunt resistance)on ITC performance was analyzed,and it was compared with the transmission characteristic of track circuit in section.(2)The transient modeling problem of ZPW-2000 series track circuit was studied,FDTD and ATP-EMTP were used to conduct transient modeling of ZPW-2000 track circuits.The method included two interface schemes: transmission lines with lumped loads at line terminations,and transmission lines including series-or shunt-connected components.The entire module was divided into two parts,transmission line part and centralized parameter network part,and the two parts were related by a controlled current source.The transmission line was solved using FDTD and the centralized parameter network was calculated in ATPEMTP.Finally,the transient analysis model of the entire ZPW-2000 track circuit was established in this way.(3)A method for intelligent diagnosis of track circuits based on ground monitoring data was studied.This method combined neural network with fuzzy logic to realize intelligent identification of the failure mode of the track circuit.The ground monitoring data is derived from the field testing and Simulink-based circuit simulation models.During the construction of the diagnostic model,the Uniform design method was used to generate the mean values of fuzzy sets,and recursive singular value decomposition was used to optimize consequent parameters to reduce the output error.8 kinds of fault data verification shows that,the presented method can achieve higher diagnosis accuracy and computational efficiency.(4)Based on on-board monitoring data,a damage detection method for electrical insulation joints of railway track circuit was studied.Firstly,the LSAE signals of the track circuit in the normal and fault conditions were solved out through the transmission model.And the influence mechanism of track circuit faults on the LSAE signals was revealed.On the basis of MF-DFA,the generalised Hurst exponents and multifractal spectra of the LSAE signals were obtained.Then the six-dimensional vectors extracted from the multifractal spectra were used as the fault features.Finally,these features were input to the extreme learning machine(ELM)to detect the disconnection fault.(5)The transmission characteristics of the track circuit when the capacitance of the compensation capacitor faults were analyzed.It was found that the "overall characteristic" and "morphological characteristic" of track circuit LSAE curve changed gradually with the compensation capacitance value.Based on this,the method of estimating the compensation capacitance by time series similarity measurement was studied.SAX was used to describe the overall characteristic of LSAE curve,and DTW was used to describe the morphological characteristic of LSAE curve.Applying the method to the estimation of the compensation capacitance,the simulated LSAE was used to construct the capacitance estimator by the similarity measure to realize the estimation of the compensation capacitance value.