The Research Of Fault Modeling And Diagnosis Method Of High-speed Train Hydraulic Shock Absorber

As an important part of the high-speed train suspension components and the dynamics system, the hydraulic damper have great effects on running safety and stability of high-speed train. Under the condition of train speed raising, the high-speed train was suffering from more severe vibration and was long-term exposed to the outdoor environment.The hydraulic damper was affected and eroded by the wind, the rain, the sun, the atmosphere and other different road conditions. It caused some fault phenomena,such as leakage. As the influence of oil, the hydraulic damper fault has strong concealment. This makes it a difficult problem,that the analysis of the influence relationship between fault and the parameters by mathematical modeling directly. What’s more, the different ways and classification standards of fault data collection caused many difficulties to the failure research.Moreover, some random excitation in complex operation condition, such as track irregularities, makes more environmental noise and mutation samples which have great influence on accuracy of fault diagnosis.Therefore, the structure and working principle of hydraulic damper were analyzed in this paper, based on which this simulation model was constructed. Through summarizing the typical failure mode and analyzes its failure mechanism, fault injection and joint dynamics simulation was realized, and the corresponding vibration signal was obtained. Based on the data, fault feature extraction methods were studied and a feature recognition method was improved. So as to put forward a set of complete fault diagnosis method of the high-speed train hydraulic damper.The main research results of this paper are as follows.(1) Fine model of hydraulic damper was built and validated. Based on the AMESim,through the modeling expression of hydraulic damper and HCD redesign of the key parts, a fine hydraulic simulation model was built, which realized the improvement of traditional simplified model of hydraulic damper in the vehicle dynamic model of SIMPack.The model was verified by compareing the indicator diagram of hydraulic damper mathematical and engineering testing model.(2) The fault simulation method of hydraulic damper was studied. Through the study of typical fault modes and the corresponding indicator diagrams, and the analysis of relationship between parts failure mode and effect, the fault trees of hydraulic damper typical failure modes were built,which verified by compareing the indicator diagram of hydraulic damper engineering testing model. Based on it, the mapping relationship between failure modes and AMESim model parameter sets was connected, and the typical faults were injected into the model.(3) VPMCD pattern recognition algorithm was improved .Based on characteristics of the high-speed train failure data,that the input data at the same stage exist mutations but was mostly continuous and stability, and the analysis of sample size requirements of VPMCD identification method in large dimension feature extraction, the original VPMCD method was improved by the form of sample set input,sliding window stepwise regression,reference associated samples and smooth processing. The improved algorithm effectively raised the recognition rate and stability of fault diagnosis. And it was verified by the high-speed train hydraulic damper dynamics simulation instance.(4) It provides a theoretical basis for monitoring point arrangement in the health status of hydraulic damper by the study of the high-speed train hydraulic damper dynamics simulation instance.At the same time, experimental analysis indicates, AS-VPMCD can apply to other complex system with sample data relevancy and continuity.