Monitoring Method Of Gear Shaper Tool Based On The Combination Of Working Condition Perception And Simulation

Gear shaping machine is an important gear production equipment.In the processing of large ring gear parts,it has much higher efficiency and flexibility than other gear manufacturing equipment.The cutting impact on the gear shaper tool is large,and the tool is very easy to wear.Therefore,the surface of the tooth is covered with a strengthened coating to improve the wear resistance of the tool.When the strengthening coating is worn out,the cutter teeth enter an abnormally worn state.With the acceleration of blutting,the cutting impact and friction are greatly increased.If abnormally worn teeth are not detected in time,the quality of the finished gears will be affected,and even serious production accidents such as tool burns and machine tool damage will be caused.In order to realize real-time monitoring of gear shaping tool wear status in an industrial environment,this paper proposes a gear shaping tool condition monitoring method based on the combination of working condition perception and simulation.The specific work is as follows:Aiming at the problem that the traditional gear shaping process monitoring dynamometer and acoustic emission acquisition equipment are difficult to integrate into the industrial environment,the physical features of gear shaping tool wear are analyzed,a measurement method combining current and vibration signals is proposed,and a sensor signal acquisition platform is built,realizing the indirect measurement of gear shaping cutting load.Aiming at the problem of complex sensor original signal components and many random interferences in the gear shaping process,a noise filtering scheme based on timespectrum entropy analysis is proposed to purify the vibration signal.The trend analysis of the pre-processed vibration and current signals is carried out,and the signal characteristics that can reflect the change of the cutting load intensity are obtained.In view of the particularity of the real-time change of the monitored object caused by the cutting process of gear shaping,a method of real-time tracking of the measured tool tooth using geometric cutting simulation is proposed,and a signal feature decomposition method based on the simulated chip volume is innovatively proposed to realize the sensor to each tooth.Time-sharing multiplexing transforms the monitored objects into mutually independent cutter teeth.The feature sequence of the cutter teeth is extracted twice,and the wear state feature vector table of the cutter teeth is obtained as the input of the identification model.Finally,the scheme proposed in this paper is verified.On the premise of conforming to the basic physical principles of tool monitoring,a reasonable supervised learning model is selected based on the physical meaning of the feature,and the abnormal wear state identification model of the tool tooth based on the Ada Boost algorithm is selected.The accuracy of the model is greater than 97% under all working conditions.Finally,based on the model,a wear evaluation experiment was performed on the entire gear shaper tool.The model's positioning results for abnormally worn teeth are all within ±1 tooth,which verifies the effectiveness of the methods in this paper.