摘要
针对刀具磨损状态分类识别精度不高的问题,提出一种基于MCADBO-SVM的刀具磨损状态监测方法。在传统蜣螂优化算法(DBO)算法基础上,引入Circle映射和自适应可变惯性权重,提出Circle自适应权重蜣螂优化(CADBO)算法,提升了算法的整体寻优和收敛性能。引入多域完全特征提取和多重特征选择技术(MFST),并将CADBO用于支持向量机(SVM)中的核函数和惩罚因子的择优问题,建立了基于MCADBO-SVM的刀具磨损状态监测模型。在公开数据集PHM2010上进行实验,结果显示:与多种方法相比,此模型的综合性能最优,检测准确率达到了95.24%。
Aiming at the problem of low accuracy of tool wear state classification and identification,a tool wear state monitoring method based on MCADBO-SVM was proposed.Based on the traditional dung beetle optimizer(DBO)algorithm,the Circle adaptive weight dung beetle optimizer(CADBO)algorithm was proposed by introducing Circle mapping and adaptive variable inertia weights.The overall optimization and convergence performance of the algorithm was improved.The multi-domain complete feature extraction and multiple feature selection technique(MFST)were introduced,and CADBO was used for the optimization of kernel function and penalty factor in support vector machines(SVM),and an optimization algorithm for tool wear monitoring based on the MCADBO-SVM algorithm was established.Experiments were conducted on the publicly available dataset PHM2010.The results show that the model has the best overall performance and the detection accuracy reaches 95.24%compared with traditional methods such as SVM.
作者
吴洪宇
徐冠华
唐波
秦炜
WU Hongyu;XU Guanhua;TANG Bo;QIN Wei(College of Metrology Measurement and Instrument,China Jiliang University,Hangzhou Zhejiang 310018,China;State Key Laboratory of Fluid Power and Mechatronic Systems,Zhejiang University,Hangzhou Zhejiang 310027,China;Key Laboratory of 3D Printing Process and Equipment of Zhejiang Province,Zhejiang University,Hangzhou Zhejiang 310027,China;Zhejiang Hangji Machine Tool Co.,Ltd.,Hangzhou Zhejiang 310012,China)
出处
《机床与液压》
北大核心
2025年第5期64-74,共11页
Machine Tool & Hydraulics
基金
国家自然科学基金青年科学基金项目(51805477)
浙江省尖兵领雁研发攻关计划(2023C01059)。
关键词
刀具磨损监测模型
振动信号
蜣螂优化算法
支持向量机
特征降维
tool wear monitoring model
vibration signal
dung beetle optimization algorithm
support vector machine
feature dimensionality reduction
