Method for predicting and diagnosing fault of a gearbox of wind turbine generator

Abstract

The invention discloses a method for predicting and diagnosing a fault of a gearbox of a wind turbine generator, which comprises the following steps of: revealing certain significant relevance among data by establishing an ARIMA (Autoregressive Integrated Moving Average)-based time sequence model for continuously changing state monitoring data, and carrying out continuation on historical data according to the time sequence model. Therefore, the operation condition of a complex mechanical system is obtained by predicting the state trend, an ARIMA model is established for historical sample data of the outlet pressure of the gearbox, the main change trend of the short-term outlet pressure state can be predicted, the prediction precision is satisfactory, and a certain reference value is provided for fault state analysis of the gearbox, due to the fact that prediction deviation of local sampling points can be large due to the change degree of single state monitoring data and the model parameter suitability degree, relevant parameters need to be corrected, and the accuracy and reliability of fault trend prediction need to be further improved together with other continuously-changing index data.

Description

technical field [0001] The invention relates to the field of fault diagnosis of wind turbines, in particular to a method for fault prediction and diagnosis of a gearbox of a wind turbine. Background technique [0002] As the key mechanical component of wind power generation equipment, the main shaft bearing of the wind turbine has high maintenance cost and long maintenance time. Abnormalities generated during the non-maintenance period are likely to cause the main shaft bearing to fail. The traditional SCADA system cannot timely and accurately locate hidden troubles. , affects the normal operation of wind turbines and the stability of wind power grid connection, so it is necessary to conduct in-depth research on the real-time operating status and fault diagnosis of wind turbine spindle bearings. [0003] The traditional spindle bearing fault diagnosis method is based on the vibration signal of the spindle bearing in the running state to carry out pattern recognition on the g...

AI Technical Summary

Problems solved by technology

However, the neural network is prone to fall into the local minimum problem, and there will be over-adaptation phenomenon.

On the other hand, the parameter optimization of the particle swarm optimization algorithm determines the convergence accuracy and prediction accuracy. The existing parameter optimization algorithms that support the particle swarm model have poor convergence speed, which greatly affects the accuracy and speed of fault diagnosis.

Images