Research On Comprehensive Evaluation Method Of Track Circuit Operation Quality

Track circuit is an important signal equipment of the railway transportation system.Its operation quality directly affects the safety and efficiency of train operation.Affected by the working environment and the aging of parts,the equipment will have unpredictable abnormalities.At present,a signal centralized monitoring system and a comprehensive track inspection vehicle dynamic detection system have been developed for track circuits to monitor their working status,and it is impossible to fully understand the operating quality of track circuits.In order to fully mine the track circuit status information reflected in the detection data,this paper designs a comprehensive evaluation method for track circuit operation quality based on multi-source information fusion,which realizes the track circuit operation quality evaluation and facilitates maintenance personnel to understand the track circuit’s working status in time.The specific research content of this thesis:(1)According to the characteristics of the comprehensive track inspection vehicle detection data,a main track signal acquisition algorithm based on frequency and variance characteristics is designed,which provides data support for the evaluation of track circuit operation quality;according to the existing literature that affects the track circuit operation quality Six of the eight indicators with technical specifications are designed with a percentile evaluation function,and a method for determining evaluation function parameters based on technical specifications and hierarchical clustering is proposed;for the two indicators of compensation capacitance and ballast bed resistance,an in-depth analysis The correlation of VMD modal components,and an improved VMD is proposed to extract the state information of compensation capacitance and ballast bed resistance.(2)For the comprehensive evaluation of track circuit operation quality,a comprehensive evaluation method of track circuit operation quality based on decision-level data fusion is proposed.This method first introduces two methods,fuzzy comprehensive evaluation and fuzzy inference,and then introduces variable weight functions and nonlinear operations to improve fuzzy comprehensive evaluation for the shortcomings of fuzzy comprehensive evaluation that will cover up "bad indicators";it is difficult for fuzzy inference rules Definite and error-prone shortcomings,a three-layer fuzzy inference model is proposed,which can effectively reduce the number of rules and simplify the traditional fuzzy inference model.Finally,based on the results of these three methods,evidence theory is used to make fusion decision.The fusion result is The running quality of the track circuit.(3)An improved method of evidence theory is proposed.This method first refines the evidence into focal elements,then calculates the focal element discount factor based on the Manhattan distance between the focal elements,and finally corrects the evidence sources and merges them.Compared with the classical evidence theory,this method can effectively deal with the conflict between the evidences,and the effect is better than that of several classical improvement methods.Using the comprehensive evaluation method proposed in this thesis,8 typical track sections are selected from actual data for analysis.The results show that the method in this paper can effectively identify the operating quality level of track circuits and has strong practicability.