Research On CNC Machine Tool Energy Consumption Prediction And Application Based On Data-Driven Method

With the increasing prominence of energy and environmental problems,energy conservation and emission reduction are important measures to be advocated and implemented in today's times.As the main production equipment of machinery manufacturing industry,CNC machine tools are numerous.It has high energy consumption,low energy efficiency and other shortcomings,and is also the key object of energy saving and emission reduction in manufacturing production and processing.Therefore,this paper takes CNC turning processing system as the research object,and the establishment and prediction method of energy consumption model in CNC machine tool machining process are studied.Firstly,the grey relativity analysis introduced to determine the main influencing factors and influence degree of energy consumption.according to the energy consumption system of CNC machine tool,the factors affecting energy consumption in each period of working condition are analyzed.The orthogonal test was carried out through the power data acquisition platform of CNC machine tool,and the experimental data were obtained.Using the grey relativity analysis to analyze the data,determine the processing process duration,material hardness,back cut,no-load power,feed speed,spindle speed,feed as the main influence factors of the energy consumption and the ranking of their influence degree.Secondly,based on the analysis of the characteristics of BP neural network algorithm and Adaboost algorithm,a modeling method of integrated optimization BP neural network based on Adaboost algorithm was proposed.The prediction model of BP-Adaboost integrated algorithm was established,and the feasibility of the model prediction was verified by the cutting case analysis.The model can improve the accuracy of CNC machine tool energy consumption prediction in the environment of small sample prediction.Finally,BP-GA and BP-PSO algorithm models were introduced into the energy consumption prediction to determine the advantages of the prediction model of BP-Adaboost integrated algorithm,and the prediction results of the three algorithm models were compared with the actual energy consumption values.The research shows that the integrated optimized BP-Adaboost algorithm model has higher accuracy in energy consumption prediction.This research not only provides a new model for predicting the energy consumption of machine tool machining process,but also provides a method support for the optimization of process parameters,and has a broad application prospect in the aspects of machine tool energy-saving processing.