Research On Welding Quality And Surface Defects Of Coffee Capsule Aluminum Foil

The production of coffee capsules is mainly composed of coffee capsule prefabrication and coffee capsule filling and encapsulation.The aluminum foil welding at the bottom of the coffee capsule is a key process in the production process of the coffee capsule prefabrication..In order to further realize the automation of the production of coffee capsules,on the basis of the design of the aluminum foil welding equipment at the bottom of the coffee capsules,the machine vision and deep learning technology were used to study the welding quality and surface defect detection methods of the coffee capsule aluminum foil.First of all,according to the functional requirements of the aluminum foil welding at the bottom of the coffee capsule,the overall structure design of the aluminum foil welding equipment at the bottom of the coffee capsule and the design and selection of its quality detection system are completed.By analyzing the problems that may occur in the actual production of the designed welding equipment,the structure of the transmission system in the designed welding equipment is optimized.Kinematics simulation and rigid-flexible coupling analysis were used to study the optimized transmission system,and the feasibility of the optimization scheme was verified by the simulation results.Secondly,through the analysis of the demand for aluminum foil welding quality inspection,the four key parameters that affect the aluminum foil welding quality,including the concentricity of the aluminum foil and the cup,the width of the welding ring,the size of the aluminum foil,and the pressure welding ratio are determined.Using machine vision technology,four detection methods for aluminum foil welding quality parameters were studied,and an image processing algorithm for detecting various parameters was designed.By comparing the test data with the actual data,the effectiveness of the aluminum foil welding quality test method is verified.Finally,in view of the shortcomings of current aluminum foil welding product surface defect detection algorithms,such as low accuracy and the need to manually extract features,a detection method of aluminum foil welding product surface defects based on deep learning is proposed.Taking five images of aluminum foil welding products as normal,dirty,wrinkle,scratch and damage as research objects,a convolutional neural network for aluminum foil defect classification was constructed,optimized and trained,and the detection effect of this detection method was verified by experiment.The experimental results show that this method can recognize 99.5% of aluminum foil surface defects,and the detection time of a single image is not more than 0.8s,which can meet the requirements for automatic defect detection of aluminum foil welded products at the bottom of coffee capsules in actual production.