The specific energy consumption of machine tools and surface roughness are important indicators for evaluating energy consumption and surface quality in processing. Accurate prediction of them is the basis for realizing processing optimization. Although tool wear is inevitable, the effect of tool wear was seldom considered in the previous prediction models for specific energy consumption of machine tools and surface roughness. In this paper, the prediction models for specific energy consumption of machine tools and surface roughness considering tool wear evolution were developed. The cutting depth, feed rate, spindle speed, and tool flank wear were featured as input variables, and the orthogonal experimental results were used as training points to establish the prediction models based on support vector regression (SVR) algorithm. The proposed models were verified with wet turning AISI 1045 steel experiments. The experimental results indicated that the improved models based on cutting parameters and tool wear have higher prediction accuracy than the prediction models only considering cutting parameters. As such, the proposed models can be significant supplements to the existing specific energy consumption of machine tools and surface roughness modeling, and may provide useful guides on the formulation of cutting parameters.
A Novel CNC Milling Energy Consumption Prediction Method Based on Program Parsing and Parallel Neural Network
