基于PCA和SVM的微铣刀磨损状态识别

Wear State Recognition of Micro Milling Tool Based on PCA and SVM

刀具磨损的状态识别精度对机床加工的质量和生产效率至关重要,为了提高刀具磨损状态的识别度,提出了一种能够有效识别微铣刀磨损状态的方法。首先对采集到的振动信号进行时域和频域分析,提取多个时域特征和频域特征,然后应用主成分分析法对提取的特征进行信息融合,再以融合后的特征向量作为支持向量机的样本输入,避免由于支持向量机初始参数选择不合适而带来的局部最优和过拟合的问题,建立遗传算法优化支持向量机的模型。结果表明,与其它算法相比,提出的基于花朵授粉算法的支持向量机模型能够有效准确识别微铣刀的各种磨损状态,优化后的支持向量机模型的总体识别率有了明显的提升,达到了95.5%,在分类性能和计算效率方面都具备优势。

The accuracy of tool wear state recognition is very important for the machining quality and production efficiency of machine tools.In order to improve the recognition degree of tool wear state, an effective method to identify the wear state of micro milling tool is proposed.Firstly, the collected vibration signals are analyzed in time domain and frequency domain, and multiple features in time domain and frequency domain are extracted.Then, principal component analysis is used to fuse the extracted features, and the fused feature vectors are used as the sample input of support vector machine to avoid the problems of local optimization and over fitting caused by inappropriate selection of initial parameters of support vector machine The model of support vector machine is optimized by genetic algorithm.The results show that, compared with other algorithms, the proposed SVM model based on flower pollination algorithm can effectively and accurately identify various wear states of micro milling tool, the overall recognition rate of the optimized SVM model has been significantly improved, reaching 95.5%,and has advantages in classification performance and computational efficiency.