Equipment fault diagnosis method based on model

Abstract

The invention relates to a fault diagnosis method based on a model, which recognizes unknown nonlinear terms by utilizing a least square support vector machine and a back scrolling method; a recognition model is added to a nonlinear proportional high-order integral observer, and thereby, the influences of the unknown nonlinear terms on fault diagnosis accuracy are eliminated, and the rapid accurate estimation of system faults is realized. The fault diagnosis method is disclosed in allusion to that the fault diagnosis problems of the unknown nonlinear terms exist in the system model; a specific equipment fault model is utilized to verify the fault diagnosis method, and the fault diagnosis method can be expanded to the fault diagnosis of other equipment which can be described by the model.

Description

technical field [0001] The invention belongs to the field of automation and relates to a technical method for equipment fault diagnosis. Background technique [0002] In recent years, with the rapid development of modern industry and science and technology, especially the development of computer technology, the structure of the process industry system is becoming more and more complex, the scale is getting bigger and bigger, and the investment is getting higher and higher. Each system of a large-scale equipment is often composed of a large number of working parts, and the different parts are related to each other and tightly coupled. On the one hand, this improves the automation level of the system and brings considerable economic benefits to production. On the other hand, due to The factors that affect the operation of the system increase suddenly, making it more and more likely to cause failure or failure. The ultimate goal of fault diagnosis technology is to avoid faults...

AI Technical Summary

Problems solved by technology

[0003] Aiming at the fault diagnosis problem of unknown nonlinear items in the system model, the present invention proposes a model-based fault diagnosis method, which uses the least squares support vector machine and rollback method to identify unknown nonline

Images