粒子群算法优化的车削温度组合预测模型研究  

Research on combined prediction model of turning temperature optimized by particle swarm optimization

作　　者：李大权 李顺才[1,2] 吴春力 Li Daquan;Li Shuncai;Wu Chunli(School of Mechanical and Electrical Engineering,Jiangsu Normal University,221116,Xuzhou,China;JSNU-SPBPU Institute of Engineering,Jiangsu Normal University,221116,Xuzhou,China)

机构地区：[1]江苏师范大学机电学院,徐州221116 [2]江苏师范大学江苏圣理工学院,徐州221116

出　　处：《应用力学学报》2020年第6期2354-2361,I0004,共9页Chinese Journal of Applied Mechanics

基　　金：国家自然科学基金项目(51675250);徐州市科技计划项目(KH17002);江苏省大学生实践创新训练计划项目(201810320024Z);江苏师范大学研究生科研创新计划项目(2019XKT159)。

摘　　要：为了研究车削温度与车削振动、车削参数的相关性,搭建了车削温度-车削振动同步测试系统。通过车削试验,利用红外测温仪及三向加速度传感器采集了车刀刀尖附近的车削温度及车削振动时域信号,分析了不同车削参数下振动加速度及车削温度特征值的变化规律。基于车削温度均值与车削参数的响应面图及等高线图分析,得到了背吃刀量是影响车削温度升高的主要因素。根据试验数据,基于指数惯性权重粒子群算法,建立了车削温升均值关于车削参数、单向振动均方根值的3个回归模型;然后综合考虑3个振动方向和3个车削参数的影响,建立了基于拟合优度赋权组合预测方法的组合模型;由3个单项模型的标准误差确定权重,建立了由3个单项预测模型及相关权重表示的车削温升均值线性组合模型。组合预测模型的相关系数为0.87,大于3个单向回归模型的相关性系数,表明组合预测方法能够较好地预测车削温升,为刀具状态监测提供理论指导。In order to study the correlation between turning temperature and turning vibration and turning parameters,a turning temperature-turning vibration synchronous test system is built.Through the turning test,the turning temperature and the turning vibration time domain signal near the tool nose are collected by an infrared thermometer and a three-way accelerometer.The variation law of vibration acceleration and turning temperature characteristic value under different turning parameters is analyzed.Based on the response surface and turning contour analysis of the turning average and turning parameters,it is found that back cutting depth is the main factor affects the temperature increase of turning.According to the experimental data,three regression models of turning temperature mean value for turning parameters and unidirectional vibration root mean square value are established based on the exponential inertia weight particle swarm optimization algorithm.Then comprehensively consider the effects of three directions of vibration and three turning parameters,and a combined model is built based on the goodness-of-fit combination forecasting method.The weights of the three single-term model are used to determine the weights,and a linear combination model of turning temperature rise means is established by three single-term prediction models and related weights.The correlation coefficient of the combined forecasting model is 0.87,which is greater than the correlation coefficient of the three one-way regression models,which indicates that the combined forecasting method can better predict the temperature rise of turning and provide theoretical guidance for tool condition monitoring.

关 键 词：车削温度 车削振动 车削参数 相关性 组合预测模型 粒子群算法 

分 类 号：TG511[金属学及工艺—金属切削加工及机床]