Application of BP Neural Network on Workpiece Edge Quality Prediction in Micro-Milling

Micro-milling is widely used in micro/nano machining. However, burrs are formed on workpiece edges. Burrs influence the workpiece edge quality seriously and must be controled. There are lots of factors that influence the formation process of burrs including cutting conditions and tool structural parameters. Burr size prediction technology can provide parameters optimization to control burrs formation actively. A BP neural network has been developed for burrs size prediction in micro-milling. The structure parameters, training epochs, error goals of the neural network are discussed and analyzed. By try and test mathods, selected network has good fitting performance and generalization capability. It is validated by experiments.