Research On Key Technology Of Automatic Rivet Assembly Based On Hole Image Recognition

With the rapid development of the aviation manufacturing industry,the manufacturing process of the aviation industry tends to be automated,and the plug pin is an important process in the automatic assembly of aircraft riveting.Due to the large number of rivets,the inserting process is particularly important in the automatic riveting process.The existing automatic nail insertion system has problems such as high transmission failure rate,low hole position recognition accuracy,and slow nail insertion speed.This paper studies a set of automatic nail insertion system for this purpose.In the nailing system,mathematical modeling and simulation are carried out for the pneumatic conveying pipeline of rivets,and the optimal geometric parameters of the conveying pipeline are obtained.The laser positioning technology used in traditional studs is changed,artificial intelligence technology is used,and the convolutional neural network is applied to the hole position visual recognition of the stud system,and an error loss function for evaluating the accuracy of the hole position is constructed.Iterative stopping algorithm program for finding the optimal solution of the network.The system greatly reduces the malfunction rate of rivet conveying and greatly improves the accuracy of hole position recognition of the system.The main research contents of the paper include:(1)According to the types and specifications of different rivets,the overall design of the stud system is carried out.The working process of the nail insertion system is analyzed,the corresponding functional requirements of the system are clarified,and on the basis of analyzing the structure of each functional module of the automatic assembly and the detection requirements of the insertion nail,the overall design of the nail insertion system is carried out.(2)Detailed design and analysis of the key components of the nail system are conducted,including the straight-slot slot hopper with reliable feeding,the rotating separator with stable operation and not easily damaged parts,and the jaws with fast ejection function.The three-dimensional model of each component and the automatic nail insertion system has been established with an integrated receiving mechanism and an electromagnetically driven micro-motion platform with precise positioning function.(3)Under the premise of ensuring efficiency,in order to solve the problem of the mutual influence of various parameters such as pipeline bending radius,inner diameter and conveying pressure in rivet transmission,this paper establishes a mathematical model for optimizing pipeline inner diameter,and the mathematics of the relationship between pipeline pressure and flow velocity model.The transmission state of the rivet in the conveying pipeline is analyzed in detail,and the calculation method of the pipe inner diameter,air velocity and air pressure required for the nail feeding process is determined,and the Fluent simulation is used to verify the correctness of the theoretical calculation value of the conveying air pressure under the optimization of the pipeline inner diameter.Obtained the relevant parameters.(4)Advanced visual inspection technology is adopted for hole position recognition.An algorithm for extracting geometric parameters of holes in riveted sheets is established.Through filtering and noise reduction and threshold segmentation algorithm preprocessing,a binary image for training is generated.The binary image corresponding to the riveted sheet and its geometric parameters are marked one by one,and a convolutional neural network hole position training model is constructed.The regression method is different from the previous classification method,and the custom error loss function and determinable coefficient are used for evaluation.In order to improve the accuracy and goodness of fit,through the research of this article,the geometric parameters of the detection hole can be obtained efficiently.(5)Thin plate hole position detection experiments are carried out.A visual inspection system for the geometric parameters of the hole position is set up,and the accuracy of the measured data in the two directions based on the space position of the thin plate is analyzed.At the same time,this paper analyzes the statistical deviation of the thin plate hole diameter.The analysis result shows that the measured error meets the measurement requirements.This paper has carried out research on the key technology of automatic assembly of rivets based on hole position image recognition,and achieved the accuracy requirements of rivet assembly between aircraft wall skin and long truss.It has important theoretical significance and practical application for the research of automatic precision assembly of rivets for covering parts value.