A state monitoring method of ball screw pair based on k-means

Abstract

The invention relates to a K-mean-based ball screw pair state monitoring method, and belongs to the field of lead screw monitoring. The method comprises the steps of collecting an original vibration signal of a ball screw pair; performing feature extraction and feature selection according to the vibration signal; standardizing eigenvalues; performing K-mean training by utilizing existing full-life-cycle lead screw historical data, and taking a simulated clustering center M as an initial agglomeration point of state monitoring; when a new sample xk is added each time, storing the new sample andfirst k-1 samples in Xk, calculating the distance between the Xk and each agglomeration point in the M, finding out the agglomeration point with the minimum distance and a label thereof, and classifying the sample into a label class; updating the agglomeration point M, namely, a mean of intra-class points; and drawing labels of classes where all samples are located, and distinguishing different states by using different colors, until equipment breaks down and the data cannot be extracted. According to the method, classification labels are used as results of state monitoring, and different labels represent different health states, so that the technical problem of data dependence of an original method is solved.

Description

technical field [0001] The invention belongs to the field of screw monitoring, in particular to a method for monitoring the state of a ball screw pair based on K-means. Background technique [0002] With the continuous improvement of equipment intelligence level and the rapid development of artificial intelligence technology, PHM system is widely used in various condition monitoring and health management of equipment. As an important part of locating equipment status changes, equipment status monitoring is crucial for subsequent fault diagnosis and equipment maintenance. The process of condition monitoring is the change process of the health state of the equipment itself. Generally, the state of the equipment is divided into three stages, health state, degradation (transition) state, and failure (fault) state. Changes in the health status often indicate that the equipment begins to degrade or fail, requiring users to maintain or replace the equipment in advance. [0003] T...

AI Technical Summary

Problems solved by technology

[0003] There are many methods for state monitoring. In general, the method is based on historical data. The results are stored by simulating a large amount of historical data, and the historical data needs to contain all state data. The data with high similarity is used as the matching result to achieve state The results of monitoring and forecasting; the state monitoring method based on historical data relies too much on the size of the database, and matches the historical data to find a high similarity as the monitoring result and prediction direction, which makes the error of the monitoring result larger, and cannot include the results of all situations

Images