Research On Fault Diagnosis And Degradation Trend Prediction Of Uniformly Distributed Tooth Surface Wear In Planetary Gearboxes

Planetary gearboxes are widely used in power transmission systems of large and complex mechanical equipment such as vehicles,aircraft,wind turbines,etc.The complex structure and the harsh working environment of low speed and heavy load lead to its high failure rate and difficult to identify.As a gradual gear fault,uniformly distributed tooth surface wear fault has significant potential hazards.From the perspective of dynamics,this dissertation conducts in-depth research on the impact of uniformly distributed tooth surface wear faults on the internal excitation of planetary gearboxes,fault dynamics modeling,early fault diagnosis,and fault degradation trend prediction.This can provide a solid theoretical foundation and effective technical means for the research of planetary gearbox fault dynamics modeling and related fault diagnosis technology.The main innovative research content of this dissertation is as follows:(1)Based on Archard equation,a model for calculating the worn thickness of uniformly distributed tooth surface wear of planetary gearboxes considering the impact of the number of planet gears is established,and the impact of uniformly distributed tooth surface wear faults on the internal excitation of planetary gearboxes is studied.From the perspective of machining,using arcs and straight lines to replace the prolate epicycloids and prolate hypocyloids formed during machining,a three-segment tooth profile equation(TSTPE)for describing tooth profile is derived based on the chord tooth thickness equation.Based on the TSTPE,a time-varying meshing stiffness(TVMS)calculation model for planetary gearbox tooth surface wear is established.Considering the effect of wear on the backlash,the worn meshing relative displacement is calculated based on the backlash function.The worn meshing error is also calculated based on the tooth profile error function.Finally,the correctness of the theoretical analysis is verified by an example calculation.(2)Based on the lumped parameter method,a dynamic model for uniformly distributed tooth surface wear fault of planetary gearboxes is established,which considers worn TVMS,worn meshing relative displacement,and worn meshing error.Analyze the acceleration,angular velocity,motion period,relative displacement between components,and other kinematic characteristics of key components of the system such as sun gear,ring gear,planet gear,and planet carrier,the differential equations of motion are established.A synthetic meshing vibration signal model considering signal transmission attenuation and planet gears passing effect is constructed.Finally,taking the planet gear uniformly distributed tooth surface wear fault as an example,by comparing and analyzing simulation signals and experimental signals,the accuracy of fault dynamics modeling and synthetic meshing vibration signal model is verified.It lays a solid simulation data foundation for fault diagnosis and fault degradation trend prediction.(3)To solve the problem of weak and difficult extraction of fault characteristics for uniformly distributed tooth surface wear of planetary gearboxes,a method for early fault diagnosis based on multi-channel optimization MCKD(MCO＿MCKD)and fault characteristics energy ratio(FCER)is proposed.Firstly,the proposed MCO＿MCKD algorithm is used to extract characteristics features.Then the fault characteristics frequency is demodulated,and band pass filtering technology is used to automatically intercept frequency-domain signals.Finally,early fault diagnosis is achieved by calculating the FCER of the signal.This dissertation takes the planet gear uniformly distributed tooth surface wear fault as an example,the robustness of the fault characteristics extraction algorithm is verified using simulation signals,and the effectiveness of the proposed early fault diagnosis method is verified through experiments.(4)Aiming at the problem that it is difficult to predict the degradation trend of uniformly distributed tooth surface wear faults in planetary gearboxes with incomplete measured data,a method for the degradation trend predicting of uniformly distributed tooth surface wear faults in planetary gearboxes based on simulation data driven and derivative dynamic frequency warping(DDFW)is proposed.Firstly,a full life cycle simulation vibration signal is obtained based on the above fault dynamics model,synthetic meshing vibration signal model,and fault characteristics extraction algorithm.Then,the proposed DDFW algorithm is used to calculate the Euclidean distance between the fault signal and the health signal to achieve fault degradation trend prediction.Finally,the planet gear uniformly distributed tooth surface wear is taken as an example,the effectiveness of the proposed method for the fault degradation trend prediction is verified through experiments.