The Research Of Milling Force Of Diamond Milling Tool For Stone Machining Based On Neural Network

With the application of the computer technology in stone carving processing, stone carving products has been a great deal of application in the daily life and has a wider application prospect at the same time. At present stone carving machine technology has become more mature. However, the use of cutting tool became a difficult problem of stone carving processing. Technological parameters of stone carving processing have been using experiential parameters. So inappropriate process parameters often lead to wear tools, even fracture, and shorten tool service life. Overlarge milling force is the direct reason of tool fracture. Nowadays more research achievements of diamond saw blade force in stone processing fields provide references in practical process. However less research of diamond cutter milling force used in stone carving hinder the rapid development and application of the stone carving technology.This paper is based on the theoretical analysis of diamond milling cutter machining stone and the stone carving processing model is established by the process analysis of the diamond cutter machining stone. It deduces tool the elements expression in the process of stone carving according to the characterization and elements definition of processing tool feature. The force situation of the single particle diamond from equal cutting volume before and after cutting is analysed, the milling force formula in stone carving by combining with the typical stone carving processing is gotten. The results show that relationship among cutting depth, feed speed and spindle speed with milling force in the stone carving is close.According to the theoretical analysis results, we built online measuring system of milling force about diamond cutter processing stone. Dynamic measurement to milling force is carried out. Single factor experiment and orthogonal experiment and tool fracture experiments of diamond cutter machining stone are accomplished by usimg this system in order to provide experimental sample for neural networkfor diamond milling force prediction and validation.We established prediction model for milling force in the process of stone carving by adopting the BP neural network and RBF neural network respectively. We realize the network design, the weights initialization and network training and simulation, etc, from toolbox correlation function in MATLAB neural network. The forecast model based on the experimental data is validated and feasibility. The model can be used to predict the milling force accurately according to the different machining parameters. The prediction accuracy of two neural networks are compared in accordance with the experimental data. The results show that using BP neural network to predict the milling force can guarantee that the predictive value of the average error is less than2%, but the single error is more volatility.In contrast, the use of RBF neural network prediction model to predict the milling forces, not only the average error is less than BP neural network but also the single error volatility is relatively stable and it is closes to the actual situation, practical,and can forecast more accurately milling force of diamond miller in stone carving processing according to the processing parameters.In the paper RBF neural network prediction model and the cutting tool fracture experiment results are combined in order to giveoptimization principle of process parameter in the course of stone carving processing and provide a reference method and basis to the actual stone carving processing in the choice of process parameters. The researching results it has certain guiding significance in the aspect reducing production cost and improving processing efficiency of the stone carving process through the optimization machining parameters.