基于拉普拉斯特征马氏距离的滚珠丝杠健康评估

发布时间：2018-03-24 22:12

本文选题：滚珠丝杠　切入点：性能衰退　出处：《机械工程学报》2017年15期

【摘要】：滚珠丝杠的性能是影响数控机床加工精度的重要因素之一。提出一种机床滚珠丝杠的性能衰退及健康状态的评估方法,该方法结合拉普拉斯特征降维与马氏距离分析模型,建立不同健康状态下传感器信号样本点在特征空间中与健康值的非线性映射关系,从而得到滚珠丝杠性能衰退程度的量化评估。通过不同健康状态的滚珠丝杠性能试验,将该方法应用于滚珠丝杠的驱动电动机速度和转矩信号,通过对传感器信号的内蕴流形及其不同健康状态下的采样特征点在内蕴特征空间中分布相关性分析,以得到量化的性能评估结果。与常见方法所得结果相比,该模型能准确地反映滚珠丝杠的性能衰退趋势,鲁棒性更好。该方法可采用数控机床自带的传感器,无须改动机床整体结构,不影响其动态加工性能,可广泛应用于工业数控机床滚珠丝杠的在线实时健康状态评估。
[Abstract]:The performance of ball screw is one of the important factors that affect the machining accuracy of NC machine tools. A method for evaluating the performance decline and health status of machine tool ball screw is presented, which combines the Laplace characteristic dimensionality reduction model and the Mahalanobis distance analysis model. The nonlinear mapping relationship between the sensor signal sample points and the health value in the feature space is established under different health conditions, and the quantitative evaluation of the performance decline degree of the ball screw is obtained, and the performance test of the ball screw with different health states is carried out. The method is applied to the speed and torque signals of the drive motor of the ball screw. The distribution correlation of the sensor signal's intrinsic manifold and the sampling feature points in different health states in the intrinsic feature space is analyzed. Compared with the results obtained by common methods, the model can accurately reflect the performance decline trend of ball screw, and its robustness is better. It does not need to change the overall structure of the machine tool and does not affect its dynamic processing performance. It can be widely used in on-line real-time health evaluation of the ball screw of industrial NC machine tools.
【作者单位】：上海交通大学机械系统与振动国家重点实验室;上海航天设备制造总厂;
【基金】：国家自然科学基金(51305258) 国家科技支撑计划(2014BAD08B01) 国家科技重大专项(2014ZX04015-021)资助项目
【分类号】：TG659

【相似文献】