Time-domain fault identifying method based on coefficient variation of regression model

Abstract

The invention relates to a time-domain fault identifying method based on coefficient variations of a regression model. The time-domain fault identifying method includes the following steps of firstly, setting a reference eigenvector, and extracting a set of a reference eigenvector formed by coefficients of the regression model; secondly, setting an eigenvector to be evaluated: extracting another set of a to-be-evaluated eigenvector formed by coefficients of the regression model according to the same way of the first step when in need of fault diagnosis of equipment; thirdly, computing difference degree between the eigenvector to be evaluated and the reference eigenvector, and indicating the difference degree as vector distance; fourthly, evaluating significance level as time limit value; and fifthly, comparing and judging equipment conditions: comparing the vector distance with the limit value, and analyzing and judging whether faults exist or not. The time-domain fault identifying method is low in computing cost and high in reliability and applicability.

Description

[0001] technical field [0002] The invention belongs to the field of fault diagnosis of marine mechanical equipment, and in particular relates to a time-domain fault identification method based on regression model coefficient changes. Background technique [0003] For a large and complex system such as a ship, if a certain key equipment cannot continue to work due to failure, it will affect the normal operation of the entire system at least, and lead to serious consequences of machine crash and death. Therefore, the failure of ship mechanical equipment Identification technology is getting more and more attention. There are many physical information used for fault identification, but the most widely used and the best engineering application effect is the identification technology based on vibration field. Among the many fault identification methods based on the vibration field, the identification method based on the time domain can identify the fault of the system only thr...

AI Technical Summary

Problems solved by technology

When the signal itself changes greatly, the prediction accuracy of the regression model is generally difficult to guarantee. At this time, the identification method using the mean value of the residual is too conservative, and there is often the possibility of misjudging the fault as normal, while the identification method using the residual variance is too strict. There are often situations where normal signals are misjudged as faults

In addition, this method has a large amount of calculation and cannot realize online real-time monitoring

However, the mechanical system on the ship has a complex environment, is greatly affected by the outside, and the signal itself changes greatly, so this method is difficult to apply to the ship's mechanical system

The second is to establish a quantitative relationship between the coefficients of the regression model and the inherent characteristics of the identification object, and then design the index to determine the maximum variation range of the coefficient based on the inherent characteristics, and then judge whether the identification object is faulty (such as Ma Gao in "Time Series Based Structural Damage However, this method is more suitable for simple structures, such as trusses and beams, and it is difficult to establish a quantitative relationship between model coefficients and inherent characteristics for ship mechanical structures, so it is difficult for engineering applications

Images