Research On Process Monitoring And Arc Length Prediction Of Plasma Arc Additive Manufacturing

Plasma arc additive manufacturing technology has the advantages of high arc stability and concentration of energy density,commonly used to manufacture large metal components.But in practical application,the arc length change is difficult to maintain because of the change of stacking height.Meanwhile,the change of process parameters will also affect the stability of the manufacturing process.Therefore,aiming at the robot automatic additive manufacturing process by the plasma arc,we design the monitoring system of plasma automatic additive manufacturing process,and predict the change of arc length in the process of additive manufacturing,so as to lay the foundation for correcting the arc length in real time.Firstly,the design and research of the monitoring system of plasma arc additive manufacturing were carried out.The electrical parameter and temperature information acquisition module had been developed.The signal conditioning circuit and protection circuit had been designed and debugged,and integrated with the mobile touch all-in-one,forming the mobile touch plasma arc additive manufacturing and monitoring hardware device.Then according to the characteristics of plasma arc additive manufacturing,the electric parameters,the temperature information collection module,data analysis,data processing,data storage and display module were worked out,forming a mobile touch type of plasma arc additive manufacturing monitoring software system.Online monitoring of a variety of parameters of plasma arc process has been realized.Then the influence of the parameters of plasma arc growth on arc length has been studied.A large number of process tests were carried out on the 3 parameters of stacking voltage,stacking current and ionized gas flow.The parameters of the test process were collected and analyzed,and the influence rule of stacking voltage,stacking current and ion gas flow on arc length was obtained,laying the foundation for the study of the arc length prediction model of artificial neural network.Then in order to realize the arc length prediction of the process,the BP neural network model was established by using Matlab software.The BP neural network model takes three characteristic parameters as the compress current,the arc voltage and the ion gas flow as the input layer,the arc length as the output layer.The arc length prediction accuracy is trained by the number of different neurons in the middle layer,and the influence of the number of neurons on the accuracy of arc length prediction is analyzed.The training results showed that when the number of middle layer neurons of the BP neural network is 7,the accuracy of the neural network is higher and the average accuracy is 89.42%.We further use the designed mobile touch plasma arc additive manufacturing monitoring system,including the established BP neural network model,to monitor the plasma arc additive manufacturing process,and verify the arc length change during the manufacturing process.The system operation results showed that the system can detect the electrical parameters and temperature parameters of the plasma arc additive manufacturing process in real time,and also predict the real-time arc length change of single channel stacking in real time.Finally,the control method of arc length was preliminarily studied.The input control model based on arc length deviation and rate of change,in particular to a neural network model for real-time prediction of the arc length as the feedback,to characterize the arc length adjustment quantity of the analog quantity to establish the basic model of double input and single output of the fuzzy controller for output.In the study,Matlab software was adopted to establish fuzzy control rule table for prediction of arc length in plasma arc manufacturing process by fuzzy processing,fuzzy inference and fuzzy processing.