Research On Health Evaluation Methods Of Gear Transmission System Based On Multidimensional Spectral Features

This article takes the actual vehicle gear transmission system as the research object,and evaluates its working state by extracting the characteristic frequency.However,there are many unrelated frequencies and noise in actual vibration signals.Therefore,feature extraction under complex background noise is always a research hotspot.And because of working conditions of the gear transmission system are complex,there are many variables,such as gears,loading conditions,measuring points,prototypes,etc.Therefore,the health evaluation of gear transmission system based on high-dimensional multivariable data is always a difficult point in research.Aiming at the above-mentioned hotspots and difficult issues,this paper constructs a complete set of health evaluation method system of gear transmission system,the main research is as follows:First,this paper proposes a parameterized feature frequency extraction method.This method can effectively suppress noise and can accurately extract the characteristic frequencies from unrelated noise.The effectiveness of the method is verified by simulation and experiment.The characteristic frequency is extracted from the three dimensions of gear meshing frequency,transmission shaft rotation frequency and rolling bearing characteristic frequency,which provides data support for subsequent research.Secondly,the difference of the prototype is analyzed based on the multi-dimensional spectral characteristics.Through the amplitude distribution of all prototypes under all measuring points at a certain frequency and their mean-standard deviation graphs,the distribution characteristics of characteristic frequencies under different machines are studied.From the microscopic perspective of the key components of the transmission system,its working status is evaluated for health.Finally,this paper uses the PCA-T~2 statistical method to reduce the dimensionality of high-dimensional data,a threshold to determine whether the system is abnormal based on statistical principles is proposed.Through the magnitude of the contribution of different characteristic frequencies,the abnormal samples can be traced back to the source of the abnormality.The comprehensive evaluation of the working state of the gear transmission system can also be realized.