Flywheel Ring Gear Of Vehicle Based On Machine Vision Research On Defect And Size Rapid Detection Technology

The flywheel ring gear is the key component in the flywheel assembly to transmit the power between the starter motor and the crankshaft.To ensure the normal start of the engine drive gear and the engine,the flywheel ring gear needs to tightly couple with the drive gear.The automated defect detection and size measurement can prevent the unqualified flywheel ring gear entering the market,and minimizing the accidents caused by the poor coupling between the flywheel ring gear and the gear,even the broken ring gear.If the flywheel ring gear leaves the factory with quality problems,there are out of the standard,it will significantly reducing the working life and performance of the gear.Therefore,the defect detection and geometric measurement of the flywheel ring gear in the industrial process is significant.The research on flywheel ring gear at domestic and abroad mainly includes the analysis of error and fracture of its inner diameter detection system,and there is still a lack of research on rapid detection of the quality of flywheel ring gear.Aiming to solve quality inspection problem of the flywheel ring gear before leaving the factory,the rapid detection of flywheel ring gear defects and size measurement are realized,and the technology research based on machine vision is carried out.The hardware system for fast detection of flywheel ring gear defects and sizes is established,and then digital image processing including image features extracting,missing and residual teeth detecting,and the measurement of size of the freewheel ring gear.The system output parameters include number of teeth,center circle diameter,addendum circle diameter,tooth root circle diameter,index circle diameter,modulus,tooth thickness,and tooth pitch.The image processing process consists of(1)image preprocessing such as image denoising and image enhancement on the flywheel ring gear image,(2)edge detection based on the improved Canny operator sub-pixel toachieve image sub-pixel edge extraction,(3)the ring gear profile and the center circle profile are obtained according to the custom feature analysis.First,the amplitude edge pair extraction algorithm at the tooth root is applied to adaptively identify the missing teeth of different types of flywheel ring gears.Second,the image difference algorithm is used to extract the tooth contour,and the connected area is extracted to analyze the number of teeth and extract the characteristics of the tooth contour.The feature extraction results are screened for residual tooth judgment.Finally,for the defect-free gear ring without missing teeth and residual teeth after defect detection,the geometric size measurement is carried out.The measurement procedure includes: finding the tooth apex through convex hull inspection,finding the tooth root point through a custom distance function,and using the least squares method to simulate Combine the tooth tip circle and the tooth root circle,then,calculating the tooth tip circle and the tooth root circle diameter respectively,further calculating the index circle and modulus of the ring gear,differencing the set of the ring profile and index circle,extracting the tooth thickness profile.By inspection the left and right end points of each tooth thickness profile the corresponding distance between each end point is calculated,and the tooth thickness and tooth pitch are figure out.For the ring gear edge detection,the traditional sub-pixel edge detection algorithm is compared,then an improved sub-pixel edge detection algorithm of the improved Canny operator is proposed.The two algorithms are compared by the change of the circle diameter in the calibration plate,and the result of standard deviation and t degree of dispersion proves the improvement of positioning in the proposed algorithm.The edge positioning of the circle diameter is more accurate,and the fluctuation is small.By designing an approach for extracting amplitude edge pairs,adaptively judging missing tooth ring gears,it ismore flexible and robust than traditional algorithms that need to set the number of ring teeth in advance based on prior knowledge.At the same time,the proposed amplitude edge pair detection algorithm,is not limited to the detection of a flywheel ring gear of a specification,and is more suitable for the actual industrial production process.Experiments show that this scheme can complete defect and size detection for the three specifications of flywheel ring gears of 170F-88,465Q-96,and 188F-110.The flywheel ring gear with defects can be quickly and automatically detected.The specific location of the residual tooth is marked,and the size of the non-defective flywheel ring gear is measured,the maximum error between the measured actual size and the theoretical size is0.0910 mm,which meets the requirements of National standard named test of the flywheel ring gear of the automobile engine.Compared with the problem of flywheel ring gear defects in manual inspection and the averaging the result of measurement,the proposed method based on machine vision inspection not only effectively reduces labor costs,but also lower inspection errors,and can improve industrial efficiency,which has high practical application value.This paper illustrates a machine vision-based flywheel ring gear defect and size detection method,and the construction of detection device.Both of them are used to complete the rapid defect detection of the flywheel ring gear with missing and residual teeth.The geometrical size parameters of gear ring are accurately measured and the system displays the final results to the user.The accuracy and reliability of system suit for real application.