Study On Intelligent Cutting Conditions Of Mould Numerical Control Machine

With the high-speed of economy and the diversification of industrial parts, it is requested more than before on die and mould manufacturing industry. For chasing steps of modern manufacturing, the die and mould manufacturing is asked more speediness and higher efficiency. In recent years, the numerical control (NC) is widely used in mechanical engineering, the quality of mould's machining is proved and the cycle of part manufacturing is shortened. So the part is updated more than before. Many of new technologies combines with NC, such as flexible manufacturing system (FMS), artificial neural network (ANN) and computer integration manufacturing system (CIMS), and so on. It makes the NC technology achieves automatization, however, the emphases is intelligent. One of the NC develop direction is the intelligent machining. In this paper, it is proposed to utilize ANN to setup mapping relations between the cutting conditions and the influential factors. And the cutting conditions are selected with the ANN.For machining, it is very important aspect in machining technology that the cutting conditions determination. The determination of cutting conditions is logical or not, it is important effect on the quality of mould, the time of machining, and the cost of manufacture. Due to the influential factors interacting each other, so the determination of cutting conditions is very difficult.As a new technology in intelligent research, the ANN is provided with the characteristic of nonlinear and distributing information. ANN can dispose the question about much input and output, on the contrary, in the traditional ways it needs to model the relationships between input and output. In this paper, it is proposed to utilize the MATLAB neural network toolbox for milling machining based on the back propagation (BP) to determine the cutting conditions intelligently.On the basis of production for milling machining, the data of milling conditions are collected to setup the BP model. In the model, the vector of input and output is confirmed. The vector of input layer is the blade material, part material and blade size. The vector of output layer is the milling velocity, feeds and depth of milling. It must confirm the number of neural in the hidden layer and the hidden layer as the BP model is setup. The BP model is trained by the data which collected from manufacture, and with the MATLAB software the function is transferred to adjust the power to train the BP neural network. The data which not training can be selected to test the capability of BP network, and the BP network can determinate the output data intelligently.Finally, the output cutting conditions are setup the parameter in UG/Manufacturing, so the real time of manufacture and the outcome of simulation is shown to us. The cutting conditions from BP neural network are applied to actual manufacturing in the workshop to machining the mould. So the milling conditions, which is selected intelligently, is validated practicability and feasibility in the machining.