A RADIAL BASIS NEURAL NETWORK FOR INTEGRATED MODELING AND OPTIMIZATION OF CNC END MILLING

摘要

This paper shows that machining process can be modeled (learned) using Radial Basis Neural Network (RBNN) and then optimized directly using the learned network. Transformation method is used to convert a constrained objective function into an unconstrained one, which is then used as the error function in the process optimization phase. The inputs to the developed simulator-optimizer consist of Spindle Speed, Feed Rate, Axial Depth of Cut, Radial Depth of Cut, and Number of Flutes, and the outputs from the model are Maximum, Minimum, Average, and Standard Deviation of the resultant force. Data collected from cutting experiments are used for model calibration and validation. The model's prediction accuracy is over 90%. On the other hand, all five input parameters are adjusted to optimum values that minimize the unit production cost subject to selected physical constraints.