1. Application Research On Long Short-Term Memory Network In Fault Diagnosis
- Author
-
Hui-Min Zhang, Bo Lin, Wei-Feng Wang, Cai-Sen Chen, and Xue-Huan Qiu
- Subjects
Artificial neural network ,business.industry ,Computer science ,020208 electrical & electronic engineering ,010401 analytical chemistry ,Pattern recognition ,02 engineering and technology ,Object (computer science) ,Fault (power engineering) ,01 natural sciences ,Convolutional neural network ,0104 chemical sciences ,Data modeling ,Support vector machine ,Identification (information) ,Recurrent neural network ,0202 electrical engineering, electronic engineering, information engineering ,Artificial intelligence ,business - Abstract
In order to improve the accuracy of fault diagnosis and accelerate the speed of fault identification, this paper proposes a fault diagnosis method based on long short-term memory (LSTM) neural network, and constructs its architecture and model framework. Taking the gear failure data as the experimental object, the performance of the model is analyzed by adjusting the number of hidden layers, the number of hidden layer neurons, the learning rate, and the training times. In addition, LSTM is compared with support vector machine (SVM), convolutional neural network (CNN) and recurrent neural network (RNN) to verify that the LSTM method has a better classification effect, and the accuracy of fault diagnosis can be increased to 99.80%.
- Published
- 2018