Research On Multi-vision Measurement And Identification System For Large Forgings

Large forging is a kind of processing method using impact force or pressure to forge parts.It is flexible and free in technology,universal in equipment and simple in tools.It is suitable for single room and small batch production and is an important means of producing large forgings.At present,the forging of large forgings in China mainly relies on manual operation to control the shape and size of the workpiece.However,high temperature,excessive volume of forgings,irregularshape,complex environment and other factors in the production process will cause the size data of various parts of workpieces in the forging process inconvenient to measure and difficult to measure accurately.The rapid and accurate measurement of the size and shape of the workpiece not only helps to improve the quality of the system and reduce the rejection rate,but also can effectively save the raw materials and greatly reduce the cost.Multi-vision measurement has the characteristics of non-contact,high efficiency,can obtain global information,etc.These advantages in measuring large forgings are obvious.It is necessary to study it deeply.In lens with infrared filter matrix camera by matching feature points out four eyes forgings three-dimensional point cloud data,using machine learning techniques to identify forgings,and then through some noise reduction method,the three-dimensional point cloud data dimension reduction process,generate the appropriate format,and then based on the related technology of machine learning,and real-time measurement of long axial forgings recognition,finally displayed on the QT(cross-platform c++GUI application development framework)development of the interface.The main tasks are as follows:1.Automatically measure the diameter of the shaft section,using 5s with a precision of 5mm,so that observers can quickly and accurately understand the size information of the forging;2.2.Use the algorithm to fit the center line of the shaft,so that workshop operators and observers can quickly and intuitively understand the shape information of the forging;3.Track and identify the flange part,intuitively and accurately show the relationship between the flange plane and the center line of the left and right axle segments.Using the principle of four-camera array and machine vision,this paper realizes fast,non-contact,multi-parameter and high-precision dynamic tracking of the workpiece,extracting effective information,and then measuring its critical size and shape deviation.The accuracy of the system was verified in the laboratory,and the diameter measurement error was 5mm.The identification and real-time diameter measurement of long-shaft forgings were completed by on-site feasibility verification.The accuracy and validity of the method are verified by experiments.Figure[53]Table[19]Reference[74]...