Predicting Energy Consumption Of Machine Tools Using Neural Network Ensemble And FOA

With the development of manufacturing industry, machine tools are playing more and more important role in production. Due to many factors such as design, operation and management, the energy consumption of the machine tool is far from normal. How to accurately forecast energy consumption of machine tools is the unavoidable question of modern green manufacturing. By using the improved FOA and neural network integration,the energy consumption prediction model has been built, which can significantly improve the accuracy and stability of prediction.First of all, through Taguchi method, the energy consumption data is studied to make out the effect of processing parameters on energy consumption. Then, the prediction model for energy consumption of machine tool is set up based on the neural network ensemble and improved FOA. Finally, the improved FOA and the energy consumption prediction model of machine tools are evaluated through the experiments.Processing parameters have significant effect on energy consumption of machine tools. The impact varies from big to small, which are cutting speed, feed and cutting depth.And processing parameters show the tendency of positive correlation with energy consumption of machine tools. In order to overcome the defects of single neural network in prediction, the neural network integration is applied to establish the energy consumption prediction model which is based on improved FOA. This model can forecast the energy consumption more accurately and steadily. At the same time, the searching capability of improved FOA in solving complex problem has made a very big improvement than before.