Vibration Signal Transfer Mechanism Of Planetary Gear System And Angular Analysis Method Of Non-stationary Signal

Condition monitoring and fault diagnosis using vibration signal effectively improve the reliability of mechanical systems,difficulties presented when these technologies apply to the planetary gears due to its complex structure,dynamic behavior and vibration signal phenomenon.Theoretical modeling and numerical simulating are the methods to predict the features of the signal relating to the fault and to explain the phenomenon of the signals Meanwhile,presented models and simulating method deviates from the true system when premised on rigid transfer path,leading to the limitations of their accuracy and feasibility.In addition,the planetary gear system often operates under non-stationary condition;extra modulations further smear the complex signal.Difficulties present on processing the signal to deliver the efficient feature for the fault diagnosisIn view of the above difficulties,it is of great theoretical significance and engineering value to study the signal transfer mechanism of planetary gear system and the signal processing method on the non-stationary signals.The main researches and results of this paper include:(1)A theoretical model is established for the signal of the planetary gears based on flexible signal transfer path.The model depicts the signal as an integral transform of the excitations;the kernel function of the transform is composed of the product of Fourier transform kernel and the two-dimensional frequency response function of the transfer path Further,the method of numerical solution of the signal is proposed based on the model.It builds up the lumped parameter model of the gear system to solve the meshing force excitation,builds up the finite element model to solve the two-dimensional frequency response function of the transfer path,and builds up the discrete model to solve the numerical vibration response signal.The analysis demonstrates that the theoretical model and numerical method exhibit high accuracy and wide feasibility,providing a theoretical basis for the existing phenomenological model while pointing out their limitations,and overcoming their limitations to a large extent(2)This paper presents the study of the signal feature and its mechanism based on the theoretical model and numerical method.This study includes normal gear,distributed gear faults,and the sun gear tooth root crack as an emphasis.It extracts a new feature of the sun gear root crack,presenting as the side band of the impulse sequence of the sun gear rotational frequency relative to that of the carrier,while the impulse series is suppressed.Therefore,the impulse series is invisible,but the existence of crack impact is indirect recognizable by identifying the side bands.On this basis,a method is proposed to identify the impact features with the octaves of the relative rotational frequency of the sun gear.It demonstrates the effectiveness of the proposed method to identify the feature.The feature repeats in the broad band,therefore,it greatly improves the certainty of the crack diagnosis(3)A stochastic optimized angular resampling method is proposed to process the non-stationary signal of the gear systems.The scheme:1)peak-detects GMF harmonic and its sideband to obtain the isogonic timing sequence mapping the time-angle relationship;2)designs the stochastic process model to optimize the isogonic timing sequence;3)Angular resamples the raw signal referencing the refined optimal isogonic timing sequence.The stochastic process evolves with displacement instead of time differing from the usual one Hence,the isogonic timing sequence is directly refined and applied to angular resampling,which simplifies scheme and improves the accuracy and overall efficiency.Viewing peak detecting as some observation,so that,the scheme establishes the observation process model and the prediction process model,while utilizes the Kalman algorithm to iterative the optimization,avoiding large matrix calculation.Errors of the scheme are studied by means of statistics,providing information to improve the scheme accuracy and modeling parameter tuning efficiency.The angular resampling shows high accuracy,robustness,exhibits the feasibility on processing the non-stationary signals,offering information with certainty to diagnosis the single and compound faults of the gear system.