Gas turbine fault prediction method based on kernel regeneration Hilbert space

Abstract

The invention discloses a gas turbine fault prediction method based on kernel regeneration Hilbert space. The gas turbine fault prediction method comprises the steps: 1, acquiring data sets of all monitoring node vectors; 2, mapping all vectors into a Hilbert space by using a kernel function, and calculating a canonical correlation coefficient between the mapped vectors; 3, calculating partial correlation coefficients of any two nodes after other nodes are given; 4, setting a threshold value to calculate a search space of the strongly related nodes; 5, using mountain climbing search in the limited space, wherein directional work is completed through a scoring function, and the causal relationship between the monitoring node and other monitoring nodes is determined; 6, obtaining a corresponding monitoring system causal structure chart before the scoring frequency exceeds a set value, and using the monitoring system causal structure chart to train a fault prediction model, wherein a fault prediction model is obtained, and the fault can be predicted more accurately.

Description

technical field [0001] The invention belongs to the field of data mining, in particular to a gas turbine failure prediction method based on nuclear regeneration Hilber space detection of partial correlation coefficients between gas turbine failures. Background technique [0002] With the development of network and technology, the data both in the society and in the enterprise has increased exponentially, and the form of data has become more and more complex. The research on how to extract useful data information from huge data is also More and more, it is obvious that these data are more commercially valuable for enterprises, such as how to find the current unit operation status and predict the future trend of unit status from the unit operation detection information of the enterprise, which can effectively avoid the loss of human and material resources of the enterprise , which greatly improves the efficiency of enterprise operation, so fault diagnosis is a topic worth stud...

AI Technical Summary

Problems solved by technology

[0006] (1) The fault prediction of gas turbine adopts the method of direct knowledge representation, which is fast, but there are few faults in the knowledge base, and it is impossible to effectively predict faults when facing new faults, which may lead to diagnostic errors;

[0007] (2) The single neural network diagnosis method cannot accurately reveal the relationship between the various components inside the gas turbine system, and can only conduct superficial analysis through data, which also brings great uncertainty to the fault prediction

Images