数控铣削能耗预测及切削参数多目标优化研究

Research on energy consumption prediction and multi-objective optimization of cutting parameters in CNC milling

为了研究数控铣床节能优化问题,首先以316L不锈钢为加工对象,设计了数控铣削实验方案,并进行了实验数据分析;然后以实验数据为样本,运用BP神经网络建立了数控机床能耗预测模型,并利用蜣螂优化算法(DBO)对BP神经网络结构进行优化,建立基于DBO-BP神经网络的数控机床能耗预测模型。通过对比优化前后两模型,选择具有更高的预测精度和稳定性的DBO-BP神经网络模型与以加工成本为目标而建立的铣削参数多目标优化模型,并运用NSGA-Ⅱ对模型求解,得到最优解集,最后运用熵权TOPSIS法对最优解集进行决策,得到最优解。通过对比优化前后比能耗和加工成本,优化后的切削参数使比能耗和加工成本分别下降了33.84%和5%。研究结果表明,优化后的切削参数更加节能和节约加工成本。

To study the energy-saving optimization of CNC milling machine,this paper firstly designs the experimental scheme of CNC milling with 316L stainless steel as the machining object,and the experimental data analysis is made.Then,the experimental data are employed as the samples,and BP neural network is applied to build the prediction model of the energy consumption of CNC machine tools,and the structure of the BP neural network is optimized by using dung-beetle-roach optimization(DBO)algorithm to build the DBO-BP neural network based on the CNC machine tool energy consumption prediction model.By comparing the two models before and after optimization,the DBO-BP neural network model with higher prediction accuracy and stability and the machining cost are selected to build a multi-objective optimization model of milling parameters,and NSGA-Ⅱis applied to solve the multi-objective optimization model of milling parameters to obtain the optimal solution set,and finally entropy right TOPSIS is applied to determine the optimal solution set.By comparing the specific energy consumption and machining cost before and after optimization,the optimized cutting parameters reduce the specific energy consumption and machining cost by 33.84%and 5%respectively.Our study shows the optimized cutting parameters achieve higher energy efficiency and save machining cost.