Quality Forecast on Finishing Machining by Free Grinding Tools Based on Genetic Neural Network

摘要

According to the current situation of finish machining by free grinding tools, GA-BP algorithm, which integrates neural network and genetic algorithm together, is applied to finish machining. A model based on genetic neural network has been built to forecast surface quality of machined parts and to choose optimized processing parameters. Finish machining by fluid magnetic grinding tools has been illustrated by comparing theoretical calculations with measured values. The result that the error is only about 5% shows that the algorithm is better. After simple modification, such algorithm could be successfully applied in other finish machining forms by free grinding tools, consequently, reasonable machining methods could be chosen.