| As a metal structural component,steel casting plays an important role in the field of power station equipment,railway locomotive,construction and engineering machinery manufacturing.The joint connection process has a huge influence on the quality of the whole product in the production and application of steel casting products,the joints are mostly connected by welding currently,due to the deep groove of steel casting,the traditional welding method is easy to generate electric arc climbing,which makes the joint quality with poor quality and less efficiency.However,the ultra-narrow gap welding method constricting arc with flux bands is suitable for the connection of steel casting because of its advantages of low line energy,high production efficiency and good joint quality,etc.Therefore,a set of automatic connection device for steel casting pipe is developed based on the ultra-narrow gap welding method constricting arc with flux bands,in order to determine the process parameters for ensuring the welding quality,the welding experiment is carried out on the experimental platform using the same process,and the quality prediction model is established by using the data from the experiment,so as to make rapid nondestructive prediction of the quality.The connection groove of steel casting pipe is relatively narrow,the traditional columnar welding torch cannot meet its use requirement,the plate welding torch and the walking device is designed combining with the ultra-narrow gap welding process,the side wall of the welding piece is used as the track for the mobile welding.The advantages of this structural design are small size,convenient installation,easy for grinding and slag removing after the welding,and it can also fit different welding morphologies.Combining with the mechanical structural design,a control system with PLC as the controller is built on the basis of the original arc welding power source,and the parameter setting and the walking mechanism control are integrated into the touch screen and the handle box of the upper computer respectively,so as to improve the parameter setting efficiency and the walking automation level of the device.For the hardware design of the control system,the design of the walking device module and the control module are mainly completely.For the aspect of the software,the parameters setting program for the touch screen of the upper computer and the control program design of the handle box are completed.The system can adjust different process parameters according to the measured groove width for adapting to the welding under different working conditions.In order to find out the process parameters suitable for the ultra-narrow gap welding and provide the data foundation for the subsequent research,the welding samples are made,the welding experiment is conducted on the welding test device using the same process,the welding current and the arc voltage are acquired synchronously in the welding process through the sensor,data acquisition card and computer software,the profile quality of the samples is evaluated and marked as well.After analyzing the data acquired in the experiment,the time series data of the arc voltage and the welding current are extracted,and the data are processed and marked correspondingly,the quality evaluation model of the ultra-narrow gap welding is established to predict the welding quality based on the deep learning theory.After the performance comparison under the same experimental conditions with fully connected neural network,SVM,fully connected neural network based on feature extraction and other models,the quality prediction model based on the time series has relatively high accuracy rate and time effectiveness,it lays a foundation for online prediction of the welding quality. |
PDF Full Text Request