| Helicopter is low-altitude aero vehicle that widely used in both military and civilian areas such as emergency rescuing,anti-terrorism,special operations and anti-submarine patrolling.It has better flight adaptability than fixed-wing aircraft since flight flexibility and harsh environment landing ability.However,transmission of helicopter could not be redundant.Working under heavy and time-varying loads,once the transmission is damaged and failed,it will lead to serious accidents.Within helicopter main transmission,planetary gear is the critical dynamic component that connects directly to the main rotor.Health of planetary gear is a vital guarantee for helicopter flight safety.However,structure of planetary gear is more complex than other gear transmissions,the movement of gear components are more complex as well.Additionally,time-varying external loads brought in by the main rotor will apply on planetary gear directly when the main rotor rotating in uneven distributed air during flight.Moreover,dynamic responses of planetary gear will be influenced by other dynamic components such as gear and shaft in the main transmission.Thus,condition monitoring and fault diagnosis of planetary gear in the main transmission is a challenge for HUMS(Health and Usage Monitoring System).Aiming at the challenges of main transmission planetary gear monitoring and fault diagnosis,research on continuous vibration separation based fault diagnosis method for planetary gear,condition indicator performance assessment and threshold optimization method were carried out using a research method that combines theoretical derivation and experimental validations.The contributions of this research work mainly include:(1)The Time-domain Synchronous Moving Average method was proposed.Focusing on the shortage of Time-domain Synchronous Average on signal periodicity and spectrum resolution,a new method named as Time-domain Synchronous Moving Average was proposed.The influences of signal overlapping on auto-correlation and power spectrum were investigated.Properties on spectrum resolution and noise gain were discussed as well.(2)The Continuous Vibration Separation method was proposed.Focusing on the shortage of vibration separation,taking the newly proposed Time-domain Synchronous Moving Average method as foundation,a new method named as Continuous Vibration Separation was proposed based on meshing regularity analysis to overcome signal periodicity problem of vibration separation.Aiming at online applications,an online algorithm of Continuous Vibration Separation was dedicated as well.(3)Aiming at two practical problems existing in the application of Continuous Vibration Separation method in helicopter engineering,Continuous Vibration Separation under limited conditions were studied.Focusing on unmeasurable rotating speed of planet carrier,a method named as tachometer mapping was proposed to estimate tachometer signal of planet carrier when applying Continuous Vibration Separation.Focusing on strong background noise interference,combined with the Minimum Entropy Deconvolution filtering method,the processing results of Continuous Vibration Separation method were further enhanced for fault detection.(4)A condition indicator performance assessment and threshold optimization method was proposed.On the basis of reviewing the commonly used condition indicators in time domain and frequency domain,the condition indicator assessment method was briefly introduced.Based on signal detection theory and ROC analysis,a condition indicator performance assessment and threshold optimization method was proposed.The method can be used for condition indicator performance assessment,as well as threshold optimization with consideration of fault detection rate and false alarm rate.To sum up,key techniques on signal processing and condition indicator assessment& optimization for helicopter HUMS were addressed in this research work.In order to overcome the problem of signal periodicity and spectral resolution of Time-domain Synchronous Average method,a Time-domain Synchronous Moving Average method was proposed.Based on the Time-domain Synchronous Moving Average method,a new method named as Continuous Vibration Separation was proposed,its online algorithm was given as well.Focusing on unmeasurable rotating speed of planet carrier,a method named as tachometer mapping was proposed to estimate tachometer signal of planet carrier.Focusing on strong background noise interference,the processing results of Continuous Vibration Separation method were further enhanced using the Minimum Entropy Deconvolution filtering for fault detection.Focusing on the selection of condition indicator and optimization of alarm threshold,a condition indicator performance assessment and threshold optimization method was proposed.The research work in this paper can provide a technical reference for the research and development of helicopter HUMS for condition monitoring and fault diagnosis of the main transmission planetary gear. |
