A continuous optimization decision making of process parameters in high-speed gear hobbing using IBPNN/DE algorithm

High-speed gear hobbing, especially dry hobbing, is an economical and environmentally friendly way to machine massive workpieces. Considering the fact that high-speed gear hobber are more expensive than the conventional machine tool and the same is true of the corresponding cutters, it is requisite that process parameters are decided appropriately so as to obtain the ideal comprehensive machining effect. What is more, the classical decision-making method of process parameters scarcely consider parameters' adjustment in the hobbing process of a batch of gears. To resolve the problems, this work presents a hybrid improved back propagation neural network/differential evolution (IBPNN/DE) approach to do a continuous optimization decision making of process parameters in high-speed gear hobbing. Gear hobbing effect evaluation model is first structured to assess the machining effect. The optimization population of the process parameters, which is the input of the DE module, is generated by the IBPNN module. Using the DE algorithm, the best of best process parameters are obtained during the course of working. Finally, a test case is presented to give a clear picture of the application of the optimization approach. The results reveal that the proposed IBPNN/DE method has better performance than the general decision-making method under the comprehensive evaluation criterion over a batch of workpieces.