基于广义S变换的磨床磨削颤振监测研究

发布时间：2018-06-09 15:25

本文选题：广义S变换 + 磨床　；参考：《浙江理工大学》2017年硕士论文

【摘要】：数控磨床是工业制造领域的常见设备,是先进制造技术的载体。颤振是导致磨床加工不稳定的重要原因之一,主要是由于刀具与工件间的相对振动引起的。颤振的发生会降低工件加工表面的加工质量、加速机床破坏。由于磨床加工过程较为复杂,且磨削颤振过程的输出信号为非平稳、非线性信号,常见的信号处理方法对其处理效果不佳。基于磨床颤振过渡信号的特点,急需引入一种有效的信号处理方法到磨床颤振信号的处理中,提取表征磨床颤振状态的特征指标,为后续磨床状态识别提供理论基础。常见的信号处理方法大多都是在傅里叶变换的基础上提出的,傅里叶变换是全局性的信号处理手段,无法表征信号的局部特性,难以分离信号的特征信息。而磨床磨削过程输出的信号为非平稳、非线性信号,在强背景信号与噪声中,更加难以识别和提取。针对以上问题,本文结合国家自然科学基金(51475432)、浙江省自然科学基金重点项目(LZ13E050003),以MK7180X160/L数控磨床为对象,采集并保存磨削过程动态变化的实验信号,利用广义S变换对实验信号进行分析,并提取表征颤振状态的特征指标,最后利用基于主成分分析法的颤振状态评估模型找出磨削特征指标与颤振状态之间的对应关系,进而提出一种实际有效的磨床磨削颤振在线检测方法。本文的主要研究内容如下:首先论述课题的研究背景和意义,介绍国内外学者针对磨削颤振机理和颤振故障监测识别领域的研究现状,并简单对比了几种常见的振动信号处理方法,指出各自存在的不足。基于磨削颤振过渡阶段信号的特点,引入一种新型的非平稳、非线性的信号处理方法-广义S变换到磨床磨削颤振信号处理与监测中,为后续研究提供方向。其次选用合适的传感器和数据采集模块,搭建实验测试平台,合理地布置振动加速度传感器,并通过改变磨床的磨削参数模拟不同的磨床状态,并实时地采集磨床的振动信号。最后将主成分分析法引入至磨床的颤振判断及识别中,建立磨床颤振状态评估识别模型,应用主成分分析法的学习能力,将实时方差和广义S变换能量熵等作为信号的特征指标,对实验数据进行了处理与分析,实现了对磨床颤振状态的判断和识别。通过对仿真信号和实验信号的处理结果,验证了本文所提出基于广义S变换和主成分分析故障诊断方法的有效性。
[Abstract]:CNC grinder is the common equipment in the field of industrial manufacturing and the carrier of advanced manufacturing technology. Flutter is one of the most important causes of machining instability in grinding machine, mainly due to the relative vibration between cutting tool and workpiece. The occurrence of flutter will reduce the machining quality of workpiece surface and accelerate the damage of machine tool. Because the grinding process is more complex and the output signal of grinding chatter process is non-stationary and nonlinear, the common signal processing method is not good. Based on the characteristics of grinding machine chatter transition signal, it is urgent to introduce an effective signal processing method to the grinding machine chatter signal processing, to extract the characteristic index of the grinding machine chatter state, and to provide the theoretical basis for the subsequent grinding machine state recognition. Most of the common signal processing methods are proposed on the basis of Fourier transform. Fourier transform is a global signal processing method, which can not represent the local characteristics of the signal, and it is difficult to separate the characteristic information of the signal. The output signal of grinding process is non-stationary and nonlinear, so it is more difficult to identify and extract the signal in strong background signal and noise. In view of the above problems, this paper combines with the National Natural Science Foundation of China 51475432and the key project of Zhejiang Natural Science Foundation, LZ13E050003, taking MK7180X160 / L CNC grinder as an object to collect and preserve the experimental signals of the dynamic change of grinding process. The experimental signal is analyzed by generalized S transform, and the characteristic index of flutter state is extracted. Finally, the corresponding relationship between grinding characteristic index and flutter state is found by using the flutter evaluation model based on principal component analysis. Furthermore, a practical and effective method for measuring grinding chatter in grinding machine is presented. The main contents of this paper are as follows: firstly, the research background and significance of the subject are discussed, and the research status of grinding chatter mechanism and chatter fault monitoring and identification field at home and abroad are introduced. Several common vibration signal processing methods are compared and their shortcomings are pointed out. Based on the characteristics of grinding chatter transition signal, a new non-stationary and nonlinear signal processing method, generalized S transform, is introduced to grinding chatter signal processing and monitoring, which provides a direction for further research. Secondly, the suitable sensor and data acquisition module are selected to set up the experimental test platform, the vibration acceleration sensor is arranged reasonably, and different grinding machine states are simulated by changing the grinding parameters, and the vibration signals of the grinding machine are collected in real time. Finally, principal component analysis (PCA) is introduced into the judgment and recognition of grinding machine chatter, and the model of grinding machine flutter evaluation and recognition is established, and the learning ability of principal component analysis (PCA) is applied. The real-time variance and the generalized S-transform energy entropy are taken as the characteristic indexes of the signal. The experimental data are processed and analyzed, and the judgement and recognition of the chatter state of the grinder are realized. The effectiveness of the proposed fault diagnosis method based on generalized S transform and principal component analysis is verified by processing the simulated and experimental signals.
【学位授予单位】：浙江理工大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TG596

【相似文献】