Method and device for predicting service life of rolling bearing based on TCN and BLS

Abstract

The embodiment of the invention discloses a method and device for predicting the service life of a rolling bearing based on TCN and BLS, relates to the field of rolling bearing fault diagnosis, and can improve the precision of predicting the service life of the rolling bearing. The method comprises the following steps: collecting vibration signals of a whole life cycle; the method comprises the following steps: preprocessing collected vibration signals to obtain a vibration signal time sequence serving as original features, and further dividing the original features to obtain a training set and a test set; and inputting the training set into a time convolutional network model, and taking the output of the time convolutional network model as the input of an enhancement node of a width learning model. And after training the width learning model through the training set, inputting the test set into the width learning model and obtaining a life prediction result.

Description

technical field [0001] The invention relates to the field of rolling bearing fault diagnosis, in particular to a rolling bearing life prediction method and device based on TCN (width learning) and BLS (time convolutional network). Background technique [0002] Rolling bearing is a mechanical element that can change sliding friction into rolling friction, which effectively reduces the consumption of friction loss. In addition, it has the advantages of easy disassembly, mass production and insensitivity to speed and load fluctuations. However, once working for a long time, rolling bearings are prone to problems, which may cause property damage at the slightest and casualties at the worst. Therefore, it is of great significance to detect the state of rolling bearings and use existing intelligent methods to predict the life of rolling bearings. [0003] Feature extraction is very important in the life prediction process for rolling bearings. Traditional methods generally use t...

AI Technical Summary

Problems solved by technology

However, in many prediction methods with deep networks, such as the common RNN (Recurrent Neural Network), LSTM (Long Short-term Memory Network), etc., the network used is usually deep, so it has a deep complexity, which leads to A major prerequisite for obtaining accurate results is the need for better hardware configuration and computing power to match the deeper network model, otherwise, forcibly using the prediction method with a deep network will make the prediction results less than expected, or the calculation time will be long too long

Therefore, in practical applications, there are problems of high hardware cost and large computing resources. To improve the accuracy of rolling bearing life prediction, more computing resources need to be occupied.

Images