Prediction Of Milling Roughness Of Aluminum Alloy Based On Tool Wear

In the cutting process,the wear state of the tool will affect the cutting performance and cutting effect,and the surface quality of the workpiece will also change.Therefore,it is necessary to consider the tool wear as an influencing factor in the surface roughness prediction model.However the traditional roughness prediction method often ignores the effect of changes in working conditions such as tool wear on the surface roughness and cutting process of the workpiece.Therefore,this paper presents a real-time surface roughness prediction method based on tool wear for aluminum alloy workpiece,using the power signal as an identification signal for tool wear.Using process parameters and tool wear obtained through monitoring,predict the surface roughness.The main research results obtained are as follows:(1)Aiming at the processing characteristics of the aluminum alloy workpiece,research on modeling nethod of tool wear monitoring model.Collect power signals during cutting and perform feature extraction.By analyzing the correlation and stability between each power characteristic value and tool wear,determine the weight coefficient of each eigenvalue to evaluate how each eigenvalue reflects the contribution of wear information.The subtraction of each wear tool and the normal tool characteristic value,and then multiply the result by the corresponding weight,add the final results,as an evaluation index ? to measure tool wear.Then using Newton interpolation establish the mapping model between the evaluation index vector and the corresponding tool wear vector,and this model is used as the monitoring model.(2)When monitoring the monitored tool,the eigenvalues regression model is established by support vector machines.The model is established to obtain the characteristic values of the normal tool and each wear tool under monitoring conditions.Then using the tool characteristic values obtained establish a monitoring model to realize the monitoring of the tool.(3)Using the BP neural network optimized by genetic algorithm,a surface roughness prediction model based on tool wear monitoring was established,namely GA-BP model.Using the process parameters and corresponding wear quantities under various working conditions,predict the roughness under corresponding working conditions.Using genetic algorithms optimized the initial weights and thresholds of the BP neural network is to achieve the purpose of optimizing the network structure,and the prediction accuracy and robustness of the model can be better than not optimized.(4)According to the modeling method of the wear monitoring model and roughness prediction model,the full-factor experimental program was designed and the model was verified by experiments.To evaluate the accuracy of the GA-BP model more accurately,this paper uses the traditional BP neural network to establish a roughness prediction model.by comparing the results of the two prediction models,it can be seen that the optimized GA-BP model has higher accuracy and better robustness.In order to analyze the roughness prediction process intuitively,a prototype system of surface roughness prediction was built using the graphical user interface GUI module in MATLAB.The system is divided into modules and each module's interface is designed.At the same time,each model's functional is introduced by using the prediction of workpiece roughness under normal tools,and verify the feasibility of the system.