基于高斯过程二元分类的刀具切削状态的识别

摘要

Based on the background of hot cell circular sawing machine cutting process, in the case of the sensor cannot be installed,this paper studied the cutting state detection and recognition problems during the cutting process. Combined with the structure characteristics of hot cell circular sawing machine, selected the servo motor torque, cutting depth and real-time cutting speed as the auxiliary variables of state recognition, based on the binary classification method of Gauss process, the cutting state recognition based on Gauss process classification is designed, after training with sample data, finally realize the classification and recog-nition of the cutting state of hot cell circular sawing machine. The identification results are compared with the actual cutting state, and the results show that, the method of cutting state of hot cell circular sawing ma-chine based on Gaussian processes binary classification can effectively and accurately identify the cutting sta-tus in the working process of hot cell circular sawing machine, and can response the actual cutting state of hot cell circular sawing machine well.%通过以热室圆盘锯床切削工艺为背景，针对热室圆盘锯床在无法安装测量传感器的情况下需对切削状态进行检测识别的问题进行了研究。通过结合热室圆盘锯床结构特征，选取伺服电机转矩、切削深度和实时切削速度作为状态识别辅助变量，在分析了高斯过程二元分类方法和锯床工艺的基础上设计了一种基于高斯过程分类的切削状态识别器，通过样本数据进行训练后，对热室圆盘锯床的切削状态进行了分类和识别。将识别结果与实际切削状态进行比较，结果表明，高斯过程二元分类的刀具切削状态识别方法能够对热室圆盘锯床工作过程中的切削状态进行有效识别，准确度较高，能够很好的反应热室圆盘锯床切削的实际状态。