目的 磨削过程中粗糙度直接影响产品质量,为有效预测工件磨削表面粗糙度,基于声发射和振动信号提出一种改进金豺算法（IGJO）优化最小二乘法支持向量（LSSVM）的磨削表面粗糙度预测方法。方法 为增强信号特征与磨削表面粗糙度相关性,利用皮尔逊相关分析和主成分分析（PCA）对信号特征进行筛选,降低特征之间的多重共线性,降低模型复杂度;为改善磨削表面粗糙度预测模型的性能,对于金豺算法（GJO）易陷入局部最优问题,在GJO基础上引入佳点集初始化种群、非线性能量因子更新策略以及融合鲸鱼优化算法改进搜索策略,提升算法的初始种群多样性、收敛精度和全局搜索能力;为提高磨削表面粗糙度预测模型有效性,利用IGJO对LSSVM进行参数寻优,建立磨削表面粗糙度预测模型。结果 通过轴承套圈内滚道磨削加工实验数据进行验证,结果表明IGJO-LSSVM磨削表面粗糙度预测模型能有效预测粗糙度值,预测精度为95.223%,RMSE值为0.013 3,MAPE值为4.776%,R2值为0.956,均优于GJO-LSSVM、LSSVM和BP神经网络模型。结论 通过IGJO优化后的LSSVM模型可实现磨削表面粗糙度有效预测,同时能够避免传统LSSVM容易陷入局部极小值的问题,对提高产品磨削质量具有重要意义。
Abstract
In the grinding process, the surface roughness directly affects the operating performance and life of the workpiece. In order to effectively predict the grinding surface roughness of the workpiece, a grinding surface roughness prediction method based on acoustic emission signals and vibration signals with the Improved Golden Jackal Optimization (IGJO) optimized Least Squares Support Vector (LSSVM) is proposed. Firstly, in order to enhance the correlation between signal features and grinding surface roughness, Pearson correlation analysis and principal component analysis (PCA) are used to screen the signal features to reduce the multiple covariance between the features and simplify the model complexity. Secondly, in order to improve the performance of the grinding surface roughness prediction model, for the Golden Jackal Optimization (GJO) is likely to fall into the local optimum problem, on the basis of the GJO, the initial population of the good point set, the nonlinear energy factor updating strategy, as well as the integration of the whale optimization algorithm are introduced to improve the search strategy, the initial population diversity of the algorithm, the convergence accuracy and the ability of the global search. After algorithmic testing, the IGJO algorithm has a higher convergence speed and global search capability compared with the GJO algorithm. After the algorithm test, the IGJO algorithm has significant improvement in convergence speed, global search ability and avoidance of local optimal solutions compared with the GJO algorithm. Again, in order to enhance the effectiveness of the grinding surface roughness prediction model and overcome the possible limitations of LSSVM in dealing with complex nonlinear relationships, IGJO is used to optimize the parameters of the LSSVM, and the grinding surface roughness prediction model is established based on the optimized model structure; And then, in order to validate the effectiveness of the above model, the grinding depth of the grinding surface is designed by setting the grinding wheel speed of machine tools, the speed of the workpiece and the grinding depth as the experimental factors. Then, in order to verify the validity of the above model, by setting the grinding wheel speed, workpiece speed and grinding depth of the machine as the experimental factors, a three-factor, three-level orthogonal experiment on the grinding of the inner raceway of the bearing ring is designed, and the vibration, acoustic emission signals and grinding surface roughness values in the process of the grinding process are collected. Finally, the above data are verified, and the results show that the use of Pearson correlation analysis and PCA to screen the signal features is effective, and the IGJO-LSSVM grinding surface roughness prediction model can effectively predict the value of grinding surface roughness, and its prediction accuracy is as high as 95.223%, the RMSE value is 0.013 3, the MAPE value is 4.776%, and the R2 value is 0.956, all of which are better than those of GJO-LSSVM, LSSVM and BP neural network models. Therefore, the LSSVM model optimized by IGJO can achieve effective prediction of grinding surface roughness, and at the same time, it can avoid the traditional LSSVM which is likely to fall into the local minima, which is of great significance to improve the quality of product grinding.
关键词
磨削表面粗糙度 /
轴承套圈 /
最小二乘法支持向量机 /
金豺算法
Key words
grinding surface roughness /
bearing rings /
least squares support vector machine /
golden jackal optimization
{{custom_sec.title}}
{{custom_sec.title}}
{{custom_sec.content}}
参考文献
[1] 王东峰, 袁巨龙, 王燕霜, 等. 轴承沟道表面完整性研究进展[J]. 中国机械工程, 2022, 33(18): 2143-2160.
WANG D F, YUAN J L, WANG Y S, et al.Research Progresses on Surface Integrity of Bearing Grooves[J]. China Mechanical Engineering, 2022, 33(18): 2143-2160.
[2] 侯智, 曾杰. 基于BP神经网络的轴承套圈沟道磨削粗糙度识别[J]. 机械设计与研究, 2019, 35(3): 119-122.
HOU Z, ZENG J.Roughness Identification of Bearing Ring Groove Grinding Based on BP Neural Network[J]. Machine Design & Research, 2019, 35(3): 119-122.
[3] 刘涛, 邓朝晖, 姚齐水, 等. 基于动态磨削过程仿真的凸轮轴高速磨削稳定性预测分析[J]. 振动与冲击, 2024, 43(10): 236-247.
LIU T, DENG Z H, YAO Q S, et al.Prediction and Analysis of Non-Circular Grinding Process Stability Based on Dynamic Grinding Process Simulation[J]. Journal of Vibration and Shock, 2024, 43(10): 236-247.
[4] 李干, 鲍岩, 王中旺, 等. 砂轮磨损检测技术研究: 现状和展望[J]. 中国科学: 技术科学, 2024, 54(7): 1263-1287.
LI G, BAO Y, WANG Z W, et al.Research on Grinding Wheel Wear Measurement Methods: Current Status and Future Perspectives[J]. Scientia Sinica (Technologica), 2024, 54(7): 1263-1287.
[5] 滕洪钊, 邓朝晖, 吕黎曙, 等. 多传感器信息融合的加工过程状态监测研究[J]. 机械工程学报, 2022, 58(6): 26-41.
TENG H Z, DENG Z H, LYU L S, et al.Research of Process Condition Monitoring Based on Multi-Sensor Information Fusion[J]. Journal of Mechanical Engineering, 2022, 58(6): 26-41.
[6] YAO Z Q, ZHANG P Y, LUO M.Extreme Learning Machine Oriented Surface Roughness Prediction at Continuous Cutting Positions Based on Monitored Acceleration[J]. Mechanical Systems and Signal Processing, 2024, 219: 111633.
[7] 迟军, 陈廉清, 杨超珍. 基于小波包分析和Elman网络的切削表面粗糙度预测方法[J]. 中国机械工程, 2010, 21(7): 822-826.
CHI J, CHEN L Q, YANG C Z.Research on Prediction of Cutting Surface Roughness Based on Wavelet Packet Analysis and Elman Network[J]. China Mechanical Engineering, 2010, 21(7): 822-826.
[8] CHEN J Y, LIN J Y, ZHANG M, et al.Predicting Surface Roughness in Turning Complex-Structured Workpieces Using Vibration-Signal-Based Gaussian Process Regression[J]. Sensors, 2024, 24(7): 2117.
[9] ARUN A, RAMESHKUMAR K, UNNIKRISHNAN D, et al.Tool Condition Monitoring of Cylindrical Grinding Process Using Acoustic Emission Sensor[J]. Materials Today: Proceedings, 2018, 5(5): 11888-11899.
[10] 尹国强, 丰艳春, 韩华超, 等. 基于声发射的磨削表面粗糙度模型及实验验证[J]. 金刚石与磨料磨具工程, 2023, 43(5): 640-648.
YIN G Q, FENG Y C, HAN H C, et al.Model and Experimental Verification of Grinding Surface Roughness Based on Acoustic Emission[J]. Diamond & Abrasives Engineering, 2023, 43(5): 640-648.
[11] 郭力, 郑良瑞, 冯浪. 基于相关性分析与CNN- BiLSTM神经网络的PSZ陶瓷磨削表面粗糙度智能预测[J]. 南京航空航天大学学报, 2023, 55(3): 401-409.
GUO L, ZHENG L R, FENG L.Intelligent Prediction of PSZ Ceramic Grinding Surface Roughness Based on Correlation Analysis and CNN-BiLSTM Neural Network[J]. Journal of Nanjing University of Aeronautics & Astronautics, 2023, 55(3): 401-409.
[12] 蔡恩磊, 王立平, 孙丽荣, 等. 基于多过程信号的轧辊磨削表面粗糙度智能预测[J]. 机电工程, 2022, 39(10): 1462-1469.
CAI E L, WANG L P, SUN L R, et al.Intelligent Prediction of Surface Roughness in Roller Grinding Based on Multi-Process Signals[J]. Journal of Mechanical & Electrical Engineering, 2022, 39(10): 1462-1469.
[13] GUO M X, ZHOU J, LI X, et al.Prediction of Surface Roughness Based on Fused Features and ISSA-DBN in Milling of Die Steel P20[J]. Scientific Reports, 2023, 13(1): 15951.
[14] PAN Y H, QIAO Y J, WANG Y H, et al.Real-Time Prediction of Grinding Surface Roughness Based on Multi-Sensor Signal Fusion[J]. The International Journal of Advanced Manufacturing Technology, 2023, 127(11): 5847-5861.
[15] 李颂华, 李祥宇, 孙健. Si3N4陶瓷轴承套圈端面磨削实验及表面质量分析[J]. 表面技术, 2021, 50(10): 363-372.
LI S H, LI X Y, SUN J.Surface Grinding Experiment and Surface Quality Analysis of Si3N4 Ceramic Bearing Ring[J]. Surface Technology, 2021, 50(10): 363-372.
[16] AXINTE D, BUTLER-SMITH P, AKGUN C, et al.On the Influence of Single Grit Micro-Geometry on Grinding Behavior of Ductile and Brittle Materials[J]. International Journal of Machine Tools and Manufacture, 2013, 74: 12-18.
[17] 任敬心, 华定安. 磨削原理[M]. 西安: 西北工业大学出版社, 1988: 206-207.
REN J X, HUA D A.Grinding Principle[M]. Xi’an: Northwestern Polytechnical University Press, 1988: 206-207.
[18] 迟玉伦, 吴耀宇, 江欢, 等. 基于模糊神经网络与主成分分析的磨削表面粗糙度在线预测[J]. 计量学报, 2022, 43(11): 1389-1397.
CHI Y L, WU Y Y, JIANG H, et al.On-Line Prediction of Grinding Surface Roughness Based on Fuzzy Neural Network and Principal Component Analysis[J]. Acta Metrologica Sinica, 2022, 43(11): 1389-1397.
[19] 张晓菊, 吴晓强, 张春友. 基于支持向量回归机的外圆纵向磨削表面粗糙度预测[J]. 机械设计与制造, 2016(8): 131-134.
ZHANG X J, WU X Q, ZHANG C Y.Surface Roughness Prediction for Cylindrical Longitudinal Grinding Based on Support Vector Regression[J]. Machinery Design & Manufacture, 2016(8): 131-134.
[20] 黄红涛, 徐文博, 李志胜, 等. 基于核极限学习机的碳化钨涂层砂带磨削表面粗糙度的研究[J]. 机械工程师, 2021(11): 137-140.
HUANG H T, XU W B, LI Z S, et al.Research on Surface Roughness of Tungsten Carbide Coating Belt Grinding Based on Kernel Extreme Learning Machine[J]. Mechanical Engineer, 2021(11): 137-140.
[21] 董浩生, 杨赫然, 孙兴伟, 等. 基于改进神经网络算法的螺杆砂带磨削表面粗糙度预测研究[J]. 表面技术, 2022, 51(4): 275-283.
DONG H S, YANG H R, SUN X W, et al.Surface Roughness Prediction of Screw Belt Grinding Based on Improved Neural Network Algorithm[J]. Surface Technology, 2022, 51(4): 275-283.
[22] 程娟娟. 高校科研与教学关系实证研究——基于皮尔逊相关系数的分析[J]. 中国高校科技, 2022(10): 46-52.
CHENG J J.An Empirical Study on the Relationship between Scientific Research and Teaching in Colleges and Universities—Based on Pearson Correlation Coefficient Analysis[J]. China University Science & Technology, 2022(10): 46-52.
[23] CHOPRA N, MOHSIN ANSARI M.Golden Jackal Optimization: A Novel Nature-Inspired Optimizer for Engineering Applications[J]. Expert Systems with Applications, 2022, 198: 116924.
[24] 谢国民, 王润良. 基于改进金豺算法的短期负荷预测[J]. 电力系统及其自动化学报, 2024, 36(3): 65-74.
XIE G M, WANG R L.Short-Term Power Load Prediction Based on Improved Golden Jackal Optimization Algorithm[J]. Proceedings of the CSU-EPSA, 2024, 36(3): 65-74.
[25] 回立川, 曹明远, 迟一璇. 融合粒子群的改进金豺算法及应用[J]. 计算机集成制造系统, 2024, 30(5): 1733-1744.
HUI L C, CAO M Y, CHI Y X.Improved Golden Jackal Algorithm Based on Particle Swarm Optimization and Its Application[J]. Computer Integrated Manufacturing Systems, 2024, 30(5): 1733-1744.
[26] MIRJALILI S, LEWIS A.The Whale Optimization Algorithm[J]. Advances in Engineering Software, 2016, 95: 51-67.
PDF(4048 KB)
渝公网安备 50010702501715号 渝ICP备15012534号-3
