Machining Parameters Prediction Research Of BTA Drilling Based On The GABP Neural Network

Deep hole machining occupies a large proportion in the field of aerospace,weapons automobiles and other mechanical manufacturing fields.But deep hole machining is also a bottleneck that hinders the development of mechanical manufacturing industry due to its special processing technology.The drilling force produced in the process of drilling,especially the thrust force and torque,is directly related to the heat generation during drilling,the wear of cutting edges,the service life of tools and the surface roughness of the holes,and the cutting parameters have a great influence on the thrust force and torque.Therefore,it is necessary to control thrust force and torque by setting appropriate cutting parameters or predict the thrust force and torque in the drilling force before cutting.In this paper,two prediction models are established by using GABP neural network to predict the cutting parameters and drilling force.It provides a new idea for the selection of cutting parameters,the control and prediction of drilling force in deep hole processing which has a certain theoretical significance and practical application value.First,Deform-3D software is used to simulate BTA drilling,and several groups of thrust force and torque data of different combinations of cutting parameters is obtained which will be provided to establish and test machining parameters prediction models.Secondly,the prediction model is established based on the GABP neural network and the simulation data to predict the cutting parameters of BTA drilling,which can reflect the relationship between bit diameters,thrust force,torque and speed,feed.Then the speed and feed predicted value is obtained by the trained model and compared with the simulation value,also,the errors between the two are calculated.The results show that GABP neural network has fast convergence speed and small prediction error,which can well learn the relationship between drilling force and cutting parameters,and the establishment of the model is very successful.Finally,six sets of different drilling forces were input into the model,and the corresponding prediction values of cutting parameters were obtained.And next,BTA drilling experiment is carried out with these cutting parameters to obtained actual drilling force,which is compared with the given drilling force to verify the validity of the prediction model.At the same time,the prediction model of the thrust force and torque is established based on the GABP neural network,and its validity is verified by the six groups of experiments.We can predict the thrust force and torque by this model if there are given cutting parameters.The research shows that the cutting parameters prediction model can estimate the speed and feed corresponding to a given the thrust force and torque,so that the thrust force and torque produced in the actual processing can reach the desired size.When the cutting parameters are given,the size of the thrust force and torque in the drilling process can also be predicted in advance by the drilling force prediction model.It has a certain theoretical reference significance and practical application value for prediction of the machining parameters in deep hole machining.