Research On The Machining Quality Prediction System For Virtual NC Turning

Cutting force is an important influencing factor of machining quality in turning. In this paper, static cutting force prediction method in engineering is researched using artificial neural networks (ANN) at first, with the help of the ANN Lenvenberg–Marquardt(LM) arithmetic provided by MATLAB software, the relationship between the quantities of training samples and the prediction accuracy of cutting force is summarized, which provides help for predicting cutting force using ANN. In cutting force experiential formulas, the formulas parameters directly influence the precision of cutting force prediction, for this reason, a new method of optimizing cutting force formulas parameters using genetic arithmetic is proposed, in the next research, the method is also applied to optimizing the parameters of exponential roughness formula.Machining of bars is an important content in turning, as research object in this paper, the machining quality prediction of bars is researched. Researches include two aspects: machining workpieces diameter errors and surface roughness. According to the three usual methods of mounting workpieces on a turning machine, the main factors influencing workpiece diameter errors are analyzed, and diameter errors prediction models for bars turning is established. Besides, the precision and appliance scope for different prediction methods of surface roughness are determined based on the cutting experiments.On the basis of analyzing the difference between physical simulation databases and geometry simulation databases, two kinds of sharing database mode, expanded database and new built database, are proposed, the hierarchical database based on the open system is built, and the prediction of workpiece diameter errors and surface roughness in turning process is realized. A new method called discrete nodes output is proposed and applied to expressing the machining workpiece diameter, further the idea and approaches about off-line compensation of diameter errors based on prediction are proposed.In order to select proper cutting tools, an intelligent cutter selection module face to different users is developed. On the basis of analyzing optimizing methods for cutting parameters selection, according to the characteristics of different machining processes, corresponding restriction functions are employed, and optimal cuttingparameters are achieved.The hierarchical architecture for virtual NC turning simulation system is proposed, on the basis of the machining process geometry simulation system for NC turning, the modules of machining quality prediction and cutting parameters optimal selection for NC turning are developed and the functions of physical simulation and geometry simulation in the virtual NC turning environment are integrated. The method and content of individuating the virtual NC turning system is proposed, finally, cutting experiments validate the reliability of the quality prediction system and the feasibility of off-line compensation for workpiece diameter errors based on the prediction system.