Fault early warning method for gearbox of wind turbine generator system

Abstract

The invention discloses a fault early warning method for a gearbox of a wind turbine generator system, and the method comprises: S1, obtaining data, i.e., obtaining data and records in an SCADA system of the wind turbine generator system needing early fault early warning of the gearbox, wherein the time span should comprise a plurality of stages of normal operation, fault formation, fault display and the like of the wind turbine generator system; S2, selecting historical data of the wind speed, the generator speed, the generator active power, the gearbox temperature and the gearbox temperature at the last moment; S3, preprocessing data, i.e., deleting abnormal data, and then performing normalization operation on the data; S4, establishing an NSET temperature prediction model by using the five parameters; S5, predicting the temperature of the gearbox by using the established NSET model to obtain a residual error between an actual value and a predicted value; and S6, determining a residual threshold, and sending out alarm information when the residual exceeds the threshold. According to the method, abnormal values in the SCADA data are eliminated by using the 3 sigma-median criterion method, the data cleaning efficiency is greatly improved, and then fault early warning of the wind turbine gearbox is realized by using the NSET method.

Description

technical field [0001] The invention relates to a fault early warning technology for a wind turbine, in particular to a fault early warning method for a gearbox of a wind turbine. Background technique [0002] There are abundant sources of operating data for wind turbines, among which the data acquisition and monitoring and control system (SCADA) equipped with the turbine itself will generate a large amount of operating data every day. Reasonable use of these information-rich SCADA data can realize fault diagnosis of wind turbines to improve The operational reliability and power quality of the unit. [0003] Nonlinear state estimation is a data-driven method, and the nonlinear state estimation modeling (NSET) method has been widely used in various aspects of the industrial field. [0004] The gearbox is a key part of the wind turbine. Its structure is complex, the design is compact, and the coupling between the components is strong. In addition, under the condition of varia...

AI Technical Summary

Problems solved by technology

[0006] The purpose of the present invention is to provide a fault early warning method for wind turbine gearboxes to solve the problem that existing fault diagnosis methods cannot predict the state of equipment in advance

Images