Wind power gear box state monitoring method based on long-short-term neural network and automatic coding machine

Abstract

The invention discloses a wind power gear box state monitoring method based on a long-short-term neural network and an automatic encoder, and the method comprises the steps: segmenting a sample through employing a sliding window in combination with a bidirectional long-short-term neural network, capturing a time dependence relation in multi-dimensional time series data, and carrying out the reconstruction mapping of multi-dimensional nonlinear time series data to a low-dimensional space through learning a large number of normal samples; on the basis, comparing and analyzing the difference between a reconstructed sample and an original sample to achieve state monitoring and fault diagnosis of the wind power gear box. The method provided by the invention can effectively process high-dimensional time series data, and can be better applied to a variable working condition operation environment of the wind power gear box.

Description

technical field [0001] The invention relates to a method for monitoring the state of a wind power gearbox based on a bidirectional long-short-time neural network automatic encoding machine. Background technique [0002] With the rapid development of information fusion technology, modern electromechanical systems can realize real-time monitoring and health assessment of operating status and processes by installing multiple types of sensors, smart meters and other equipment. With the rapid development of machine learning and deep learning technology, data-driven models have become an effective method for fault detection of electromechanical equipment, especially wind power gearboxes. Considering that the data collected during the operation of the wind power gearbox is multi-dimensional time-series data, it is very important to perform condition monitoring and fault diagnosis on the wind power gearbox based on the multi-dimensional time-series data. However, the processing and...

AI Technical Summary

Problems solved by technology

However, the processing and analysis of multi-dimensional time series data is difficult: First, there are potential correlations and mutual influences between different dimensions of the data, which lead to the fault detection and identification of wind power gearboxes.

Secondly, the wind power gearbox is always in the non-stable operating condition of variable speed and torque, and its data volume is large, multi-source heterogeneous and dynamic, etc., making big data processing more difficult

Images