predicted value is the thermal model prediction values baseon the temperature coefficient combination of the desiredorthogonal experiment data and the machine temperature. The residuals are the difference between the experiment data and the predicted value. It is obvious that the smaller is the variance sum Ps, the more precisely the model predicts. According to orthogonal regression design theory, 25 orthogonal design points are needed for computing 25 objective function values. Considering the multi-condition of the working status and environment, more sets of actual cuttingdata acquired at different times should be chosen. After obtaining the objective function valueY, the following equation can be computed though the least mean squaremethod using each of the coding factors (X1,X2,X3, and X4)
đang được dịch, vui lòng đợi..
