Studies On Milling Parameters Optimization Based On Artificial Neural Network

In recent years, the emergence of large composite numerical control machining center has greatly reduced the production time of numerical control processing and increased the processing efficiency. Under the premise of ensuring the quality of the product, choosing the corrected and reasonable cutting parameters to short the cutting time, reduce the production costs and improve the production efficiency is the main target of modern machining. However, during the machining process, there are many factors for influencing cutting time, such as cutting speed, feed rate, back cutting depth, cutting width, tool, fixture, form accuracy, surface roughness and so on. Therefore, this paper use theory of modern cutting, establish CNC milling machining parameter optimization mathematical model, and use mathematical method and simulation software to optimize the processing parameters and seek the optimal combination of cutting parameters.Through the study of 3D AC servo closed loop NC Worktable machine structure, motion path and the machining parameters,making dashboard parts as the processing object, in three-dimensional modeling software UG to build machine, instrument panel model. Based on the simulation platform of VERICUT, the virtual machining system of machine tool is built, and the reliable data of the instrument panel is obtained by the method of simulation and experiment.Based on machining simulation and actual machining to obtain the data, to artificial neural network training, and carries on the forecast to the network, a multi-objective optimization on the output of the network using genetic algorithm, we get the value of the optimal processing parameters in the dashboardThis paper is based on the 3D AC servo closed-loop NC workbench which is designed and developed by solid Tech(Shenzhen) Gu Gao, and the entity processing of the instrument type parts. The results show that the optimized cutting parameters can be shortened effectively by using artificial neural network, and the accuracy and reliability of the optimized scheme are verified.