Research On Visual Measurement Method Of Medium And Small Module Gears And Tooth Pitch Measurement Technology

Gears are standardized typical transmission parts,which have been widely used in machinery,aerospace,instrumentation,military,and other fields.Gear transmission accuracy directly affects the transmission power,working accuracy,load-bearing capacity,noise and service life of the machine.In the process of gear processing,gear detection is an important means to evaluate the quality of gears and is an effective way to improve the technological level of gear processing.With the continuous improvement of machine vision technology and computer image processing technology,gear measurement technology based on machine vision has also been developed rapidly.Under the support of the National Science and Technology Support Program named Precision Mechanical Transmission Design and High-grade Gear Manufacturing Technology（2014BAF08B01）,thegearmeasurementtechnologyofsmalland medium-modulus for high-precision based on the partial visual image of the gear is studied deeply in our research group,and the vision measurement instrument of small and medium-modulus spur gears is developed.Then on the foundation,the correlated error items and the pitch measuring technology of small and medium-modulus gears are deeply studied,which is aimed to achieve the purpose of quickly measuring the small and medium-modulus involute spur gear with the 5-level precision standard.The main work and research results of the paper include:（1）The preprocessing technology of data processing image is studied.The work inciudes image filtering and denoising,gray-scale segmentation and radial segmentation of measurement area in the tooth profile.The pixel elements in the transition band of tooth profile edge are extracted.The whole parameter database of the pixel elements in the transition band of tooth profile edge is established through coordinate transformation and calculation of involute parameters,which realizes the standardization of data exchange among various measurement algorithms,simplifies the algorithm and improves the calculation speed.（2）The sub-pixel edge detection algorithm based on Bertrand gray-scale surface model is improved to improve the calculation speed of the algorithm and adapt to the needs of real-time fast measurement.The concept of the pixel elements normal in the transition band of tooth profile edge is proposed,and the concept of phase angle of the tangent point in the base circle is put forward.The recursive formula of the involute along the tangential arc length segment is established.It is used for segmentation processing of Bertrand gray-scale surface in the transition band of tooth profile edge and transforming the complex geometric quantity calculation problem into a simple generation quantity classification problem.The sub-pixel edge detection speed is greatly improved and the improved algorithm is 23 times faster than the original algorithm.The discrimination and correction of distortion for all tooth profiles by sub-pixel edge detection results is achieved in the real-time measurement process.The accurate position positioning parameter of the tooth base circle（?₀^*）is given,and the speed index of measuring one pitch in 0.7s is achieved.（3）The high-precision positioning algorithm for the base circle position of tooth profile is studied.Using the normal offset of sub-pixel edge points（?_i）and the radial segmented average normal offset of edge transition band（_it）,the iterative approximation method is respectively used to separate tooth profile slope deviation.The edge distortion of tooth profile image is discriminated by proximity discriminantion method.The position positioning parameter of the tooth base circle（?₀^*）is corrected by distortion elimination algorithm and maximum probability interval estimation algorithm of edge position,and the position positioning accuracy of the tooth base circle is improved.It has been experimentally verified that the measured pitch after the correction for the edge distortion of the tooth profile image is very close to that measured after the strict cleaning process to eliminate the edge distortion of the tooth profile image,and the deviation is less than 0.75μm.The requirement of gear cleaning and measuring operation is reduced,and the practicability of vision measuring instrument is improved.（4）Using the normal offset of sub-pixel edge elemenys（?_i）as measurement parameters,the measurement models of total tooth profile deviation（F_α）,tooth profile shape deviation(f_fα)and tooth profile inclination deviation(f_Hα)were established.The undistorted tooth profile edge is selected as the tooth profile deviation measurement object,which can eliminate the measurement error of the tooth profile deviation caused by the tooth profile edge distortion,and simplify measurement calculation of the tooth profile deviation and improve the measurement speed.（5）Based on the optimal estimation parameters for base circle position positioning of the tooth profile,the measurement models of single tooth pitch deviation(f_pt),tooth pitch cumulative deviation(F_pk),total tooth pitch cumulative deviation（F_p）,radial tooth thickness(s_rj),and the variation model of common normal length（（35）F_w）are established.The measurement calculation of the tooth profile position deviation is simplified,and the measurement speed is improved.（6）The influence law of base circle positioning eccentricity on the accuracy of tooth pitch measurement is analyzed,and the sinusoidal curve model and the incremental model of tooth pitch measurement error for the influence of the base circle positioning eccentricity on the tooth profile positioning error are proposed.It has been experimentally verified that the calculation accuracy of the incremental model of tooth pitch measurement error is high,which can be used for the precision analysis for the development of gear vision measuring instruments.It is concluded by the model analysis that improving the accuracy of gear rotation positioning in gear vision measuring instruments is an effective way to improve the accuracy of tooth pitch measurement.The measurement requirements of the 5-level precision gear can be meet when the eccentricity is less than or equal to 40 microns（e?40?m）and the positioning error is less than or equal to 1 degree（（35）?_j?1^?）.The double tooth pitch measurement method can be used to improve the efficiency of visual measurements When the number of gear teeth is greater than or equal to 45,and the measurement requirements of the 5-level precision gear can be meet.The measurement error of tooth profile shape can be affected with the increase of the eccentricity in base circle positioning,and it thus has a certain impact on the accuracy of visual measurement for pitch deviation.（7）Based on the hardware design and pre-software development of the spur gear visual measuring instrument,the core measurement software is upgraded comprehensively by using the measurement model and algorithm studied in the paper.The upgrade software adopts modular design.The system has the advantages of simple operation,fast measurement speed,high accuracy and perfect functions.The total measuring time of 90-tooth gear is less than200 seconds,and the average measuring time of each pitch tooth is less than 0.7 seconds.The accumulated pitch deviation is less than 3μm compared with the measurement results of M&M3525 gear measuring center.The total deviation measurement result of the tooth profile is less than 1μm compared with the measurement results of the M&M3525 gear measurement center.